We need to think about our priorities here. The end goal is to have a global taxonomy that can support both casual and expert use, including conservation work. Cleaning up unknown relationships is a good thing, but a usable taxonomy is usable whether or not we’ve carefully mapped it to POWO.
I definitely believe in a dynamic taxonomy, but it’s really important to be judicious about taxon changes. Poorly founded lumping can wind up destroying a lot of information that observers and identifiers have carefully entered. Even if the swap is fully 1-to-1, making a swap that later has to be reversed can litter a large number of observations with unnecessary reidentifications. Splitting a genus may require recruiting identifiers to go in and do manual reidentifications, and so on.
Frankly, I think the increasing insistence on conforming to POWO has been counterproductive in terms of generating a stable taxonomy. In theory, it makes sense: when our treatment doesn’t match POWO’s, we should be feeding back to them to see if it’s an oversight or a deliberate judgment on their part. In practice, what seems to have happened is that some curators feel empowered to charge in and swap “because POWO”, without examining relevant literature or inquiring why the deviant taxon was created in the first place.
Turning curation into a race to manually align with POWO isn’t a good use of the collective taxonomic expertise that’s accumulated at the site. I think we should be thinking more about tools and initiatives to raise awareness of what taxonomic changes are pending (being notified after the change is committed is not very helpful) and help people get an idea of what problems exist within groups of reasonable size.
I thought a little data might be useful - I just looked at the top-50 research-grade plant species in the United States. In regard to subspecies/varieties, this is what I see:
31 out of 50 species have the subspecies/varieties matching POWO. Yay! (Of those, 16 have no subspecies/varieties).
4 out of 50 species are missing subspecies/varieties, with no flags or formal deviations. Probably just a manpower problem. Not ideal, but okay.
4 out of 50 species have extra subspecies/varieties, with no flags or formal deviations. Probably a manpower/historical problem. Not ideal, but okay.
3 out of 50 species have some weird but pretty simple things going on. They don’t seem to match POWO, but I didn’t fully review.
2 out of 50 species are flagged and have a healthy ongoing dispute/discussion. Good enough for me!
2 out of 50 species are a hot mess with a million subspecies/varieties. They probably don’t match POWO, but I didn’t fully review.
1 species flagged 2018.04 and 2024.03. Extra subspecies/varieties with no formal deviation; flags have no substantive discussion.
1 species flagged 2023.09. Extra varieties have a formal deviation with no history/discussion. Flag has some good discussion, but no expressed scientific rationale for the deviation.
1 species flagged 2023.11. Extra subspecies/varieties with no formal deviation. Some discussion, but no substantive reason provided in support of the deviation.
1 species flagged 2023.12. Extra subspecies/varieties with no formal deviation. No substantive discussion.
Obviously this is a small sample size. But I suspect it’s decently representative. All this is just increasing my resolve to apply to be a curator and try to get some stuff done without getting yelled at too much. So, I’m glad we’re having this discussion.
There are a few other options. See The big four of plant taxonomy – a comparison of global checklists of vascular plant names by David Schellenberger Costa et al. – https://doi.org/10.1111/nph.18961.
The problem with the other non-POWO, non-WFO options is they’re essentially the same as POWO in that all taxonomic changes ultimately flow through (or under the coordination of) one guy. WFO makes the process more community/consensus-oriented, but is still kind of a work in progress. It possibly might be mature enough to switch to even now though, if they’ve imported all of POWO already and now are just trying to set up TENs. That doesn’t really solve the underlying problem though. Read on below.
This. It’d be nice if there were a set of principles that could be pointed to to guide the “POWO changed something, what do we do?” decision-making process. It feels like there’s an awful lot of spinning and reinventing of the wheel everytime there’s a taxon flag on a plant taxa.
The curator guide says that the site follows taxonomic authorities to minimize the need for taxonomic debates, and yet in practice “because POWO” doesn’t seem to be considered enough of a justification to make a taxon change. So, why do we follow taxonomic authorities then? Don’t say its to minimize taxonomic debates…
I get that we need to balance all of this with pragmatism, but it seems that balance isn’t really there. It should be “POWO says this so we’re thinking of changing this taxa … anyone have a reason why this is an absolutely terrible idea (in terms of practical impact on the site or its core use cases not in terms of ideology or philosophy) or can anyone point to a documented curation principle that dictates we should think about doing this differently here… no? … okay, we’re gonna do it then”.
iNaturalist only follows COL down to phyla. So the “fault” of still having Magnoliopsida in the taxonomy is purely our own. There’s a flag discussing why we’ve not fixed that yet: https://www.inaturalist.org/flags/240790.
World Flora Online is working to establish a higher-level taxonomy that’s actually monophyletic and get COL updated accordingly. See page 15 of https://about.worldfloraonline.org/images/uploads/documents/21st-WFO-Council-Final-Report-November-2023-virtual.pdf. I believe the first thing they did was recognize sub-kingdom Bryobiotina. I created a flag for that (https://www.inaturalist.org/flags/680883) though it hit the same wall as Magnoliopsida (as well as another, apparently subkingdom doesn’t exist in the codebase). They’ve also recognized Pteridobiotina but that hasn’t been accepted by COL yet. They plan to do more work at the high-level and eventually do something about the angiosperms.
I’m not even going to touch protists and chromists because that makes this an even bigger monster of a conversation.
Honestly, how does deliberately ignoring the taxonomic source creates a more stable taxonomy?
And this quote exemplifies my thoughts. As a curator, I primarily focus on animals (somewhat due to backlash I’ve received from committing several plant swaps), but imagine what would happen if we deviated from something like birds. There are plenty of things I wish we could implement with birds, but the result would be a deviation.
The point I’m trying to make here is, just because you don’t agree with the methods or taxonomic source, doesn’t necessarily justify a deviation. We have the curator guidelines for a reason, so why do plant taxonomists get the privilege to bend so many of the rules?
I’m not well-versed enough in plant taxonomy and how it works to have a valuable opinion, but I did want to chime in to ask that the thread maintain its focus on following POWO or not as opposed to discussion of CoL (a different taxonomic source) and “deep” phylogeny on iNat.
I agree that it would be nice to have a deep phylogeny that more closely matches our current understanding of Life, but this is almost entirely a separate issue to the main focus of this thread. Making changes near the base of the Tree of Life is very intensive for iNat (due to the number of observations it involves) and has very little effect on almost every user’s day to day interaction with iNat. Very few IDs are made at the broadest levels of life.
If there’s the need for a discussion about CoL and the base of the tree, that can be made its own thread (and I’m happy to move posts to a new one if that is desired) or an existing thread can be reopened.
Sure thing. I merely responded to the original point about COL to point out that it’s not used for as much of the tree as some might think. The taxonomic authorities for plants are (as I understand it):
Land plants
Catalog of Life (as of 2019) down to phyla (land plants only)
No taxonomic authority between phyla and family
POWO for family and below (vascular plants only)
FNA for family and below (bryophytes s.l. only)
Algae
AlgaeBase
I sense that raymie’s point is more immediately about our adherence to POWO for family and below and whether we’re actually doing that like we say that we do. But I do think it’s at least fair to at least point out that there is a portion of the plant taxonomy not mapped to a taxonomic authority at all, and that that, as well as having to juggle multiple authorities perhaps does exacerbate the problem of an overabundance of subjectivity and “how does this make you feel” when it comes to dealing with plant taxon flags, because at some levels of the taxonomy that’s all we can do.
WFO does provide a taxonomy for land plants from kingdom on down to infraspecies. However, even if we switched from POWO to WFO (which I’m not necessarily here proposing, and I realize is beyond the immediate scope of this topic), resulting in (hopefully) better, more consistent, and more consensus-based coverage with fewer instances where we have to figure it out on our own, we’d still have the same problem to deal with. The problem to be solved is that there are differing opinions on the criteria for when we go ahead and accept what POWO (or whatever taxonomic authority we follow) did versus when we actually need to talk about it further, specifically in the realm of plants.
I know iNaturalist isn’t supposed to be a taxonomic authority, and the direction we’ve been given in the past is that iNat should always follow outside authorities, but something that isn’t talked about enough is that there are virtually no other platforms where the world’s taxonomic experts come together and discuss taxonomy in real time. iNaturalist might be the best possible platform for maintaining an ever-evolving taxonomy. If the documentation was better, it could be argued that we are the most accurate taxonomic authority for plants.
I only wish authors were displayed with taxon names and that primary literature was referenced more often with taxon changes.
In theory, it is nice to use an up-to-date taxonomy but the problem is too complex to solve it by automatically following POWO or any reference. Let’s try to dissect it into manageable issues!
There are different forms of changes to be made to iNat taxonomy to approach POWO.
1-to-1 naming issues: this is the simplest. One old name passes its biological content to a new name in perfect agreement. These are free to implement anytime, people will get used to the new name.
1-to-N split: the meaning of an old taxon is divided between new ones (or the old one with a narrower meaning and a new one). Here one old name either stays or is replaced by a new name with a narrower meaning, and some of its observations are re-assigned to a new taxon. The crucial step here is the introduction of a new taxon to which some observations will be re-assigned. This can be committed anytime and is easy to revert by merging, see 3.
N-to-1 merging: One new name inherits the content of several older names. This is dangerous, because committing such a change can irreversibly lose the information content of the variation between the N old taxa. This may not be important at present but it can happen in the future that the N-taxa view proves true again. In this case, observations merged earlier must be re-identified one by one. As a potential solution, I strongly suggest introducing N infra-taxa in 1-to-1 correspondence with the N old taxa. These infra-taxa need not be at any official level, if there is no such candidate; however, the old taxa often remain separated at lower taxonomical level, e.g. old species become subspecies, etc. We may even introduce an arbitrary taxonomical level for this. The most important is to achieve 1-to-1 swaps between old and new units, that warrant reversibility. Of course, this works only if the folks use the infrataxa during identifications. In general, giving IDs at levels besides species and genus (e.g. sections, complexes, subspecies, etc) can ease re-identifications after the reversion of a lump.
Finally, I strongly discourage anyone from committing lumps (case 3) by uncritically following POWO, without knowing the taxonomical problem more deeply. Especially, never commit any changes to unfamiliar taxa distributed in other parts of the world without knowing them! Always assume that the seemingly outdated taxonomical view on iNaturalist is the up-to-date knowledge of scientists of an older time, and might turn valid again in the future.
agreed, I’ve seen too many instances where action ended up making things more confusing or annoying more people (and I am one of those annoyed people, I’m still not happy someone went through and made Silphium conform to POWO’s outdated treatment last year).
Plant taxonomy is a lot messier and less stable than bird taxonomy, so that’s why it is the way it is. There’s a lot fewer people working on a lot more species. It’s an issue with many groups of insects as well, but we don’t have a single overall database for all insects like we are attempting to do with plants. We also don’t have something like the Clements checkist for birds that is well-curated and relatively frequently updated. With plants it’s very haphazard in general, our taxonomic issues on iNaturalist are representative of broader issues within botanical taxonomy as a whole. Plants hybridize like crazy when compared to animals so definining what is and isn’t a species can be a lot harder.
Some suggested guidelines regarding found deviations from POWO:
If a deviation is found without justification given, first look for the justification elsewhere such as flags at the genus-level or a related inactived taxon. If nothing is found, then review the literature and/or contact experts. The literature is often behind a paywall but there are many people on iNat that can get around the paywalls and share the literature. You can also often access the literature through a local library, though you may have to access it from a physical computer at the library.
If the deviation seems justified, contact POWO and ask why their taxonomy doesn’t align with the deviation. They will often give you a reason that you overlooked, which may be a taxonomic code violation not pointed out in the literature. Alternately, they will may make the change to POWO fairly quickly or tell you that they plan to make the change to POWO but it may take awhile. They have a huge backlog, which makes a lot of sense considering that they have way too few people to do such a massive amount of work. This is why you should do as much work as possible to resolve the question before contacting them. If POWO plans to deviate and iNat has already deviated, there is nothing wrong with iNat being a little ahead of POWO.
If the deviation does seem wrong, it may be worth reverting to what POWO has but caution should be advised. As previously mentioned, any sort of lumping of multiple taxa into one is highly destructive to the data. Thousands of hours of work could be lost. Also, while there are occasionally good reasons for lumping, a lot of the current taxonomic work is trying to fix all the damage done by people who lumped taxa in the past that should not have been. It is especially good to check if any POWO lumped taxa have a rare plant status. Those are regionally important and are often the taxa that are lumped when they should not be.
If things are ambiguous because there are multiple conflicting treatments, it may be good to wait until the dust settles before making changes, though it is generally better to recognize more taxa than less for the reasons mentioned in 3. A good example of a problematic situation is the genus Amsinckia, which I found three recent treatments for that all contradict each other. Another is the mistletoe genera Phoradendron and Arcuethobium. There are some serious lumper vs. splitter disagreements in those genera and it seems that neither group has provided solid enough evidence to adequately dispute the other.
I do feel this discussion is important to have, and it is one primarily meant for curators. I have been considering applying to be a curator for quite some time.
For clarification of the rules, would this exact topic have been better in the “general” category? It’s certainly more relevant to the average curator than to the average user.
One quick reminder that a friend gave me- POWO is an attempt to standardize an entire taxonomic kingdom. That is a lot to try to deal with, so while they’re good at what they do there’s no expectation of perfection and thus deviations will be necessary to use on this website probably forever.
These are great guidelines! In particular I recommend the advice of doing your own literature research and then contacting POWO if you feel their taxonomy could be improved.
I’ve been impressed by the speed and willingness of POWO to change their taxonomy at my request, sometimes based on fairly obscure details buried in century-old treatments.
POWO did disagree with my viewpoint on a couple of requests, both related to lumping or splitting genera. Fortunately, in that scenario we’re not really losing any information by following POWO’s approach (unlike with lumping species or other “leaf” taxa), so the stakes are lower.
In one of those cases (Agavoideae), I and other iNat curators chose to deviate from POWO to retain the genera Manfreda, Polianthes and Prochnyanthes and (temporarily) accept the paraphyly of Agave. By doing this, we chose an approach that better accorded with morphological distinctions, historical literature and the understanding of field researchers. Subsequently, three new genera have been published to accommodate basal former Agave species— Paleoagave, Paraagave and Echinoagave—allowing a monophyletic Agave s.s. and avoiding the need to subsume long-recognized monophyletic genera such as Manfreda and Polianthes into Agave. So, I’m glad that we did choose to deviate and avoid adopting an unhelpful taxonomy for a couple years only to have to revert most of it later.
In my experience, deviations from POWO have been useful in two circumstances:
As above, when POWO has adopted an approach of lumping genera that doesn’t have consensus from researchers in the area and would be disruptive to iNat identifications. This definitely needs discussion via a flag with all interested parties on iNat.
When uncontroversial new taxa have been published (often supported by iNat observations) and we prefer not to wait for roughly 1 year for these new names to filter through IPNI and POWO to become available for use on iNat. I’d like to see curators explicitly note the reason for this deviation in the taxon history.
If the taxa in question are endemic to a less researched region/group, with very little recent literature, local authoritative works (and/or experts) should be consulted; in many parts outside of immediate sphere of western science (eastern Europe is far enough) there is knowledge, but its in gray or black literature, or not published at all.
I think it would’ve been best to start there, then it could have been moved here. Not a huge deal, but I wanted to just make sure the guidelines were clear.
Anyhoo, I don’t want to further distract from the actual conversation.
absolutely agree with adding these and deviating, especially imo for conservation-significant taxa. Also just to note, when I do add these species to iNat, I also go ahead and add the names to IPNI myself to kickstart the journey to POWO faster
I disagree that every flag devolves into total disagreement. I have seen curators come to compromises before.
Many actual researchers are active curators on iNaturalist. To me, it seems ridiculous to dogmatically follow POWO online when one of the world’s foremost experts in a specific genus is disagrees with their assessment and has literature to back it up. I am not trying to say that the POWO staff aren’t knowledgeable (I’m sure they are), but no one can be an expert in more than a handful of genera in their lifetime.
