Plant Taxonomy - Are We Really Following POWO? Should we?

According to the iNaturalist Plant Working Group page, we are following POWO taxonomy for vascular plants. However, in practice we aren’t. Since deviations are allowed, every time a taxonomic change happens in POWO an argument breaks out over on iNat about whether we should follow it or not. These taxonomic changes often don’t happen at all because there are too many disagreements about what to do about these, which ultimately means the taxon changes never get committed or added to the deviation list, which then causes all sorts of issues.

I see a few options on how to fix these problems:

  1. We choose to follow POWO to the letter. This results in no more augments on if taxonomic changes need to be carried out, but may result in a little grumbling (more on that later).

  2. We decide on a small, set number of deviations. This list would never (or at the very least be very, very infrequently) get updated, and follow POWO would be the default state until staff say otherwise. This is theoretically what we are doing right now, but it isn’t in practice, so I’m not sure this option is feasible, as attractive as it may sound.

  3. We decide on a different taxonomy and strictly follow that. There is already a very vocal minority of users who do not like using POWO as a reference and would strongly like us to stop using it altogether. This would require us to find a new taxonomy and commit to a very large number of taxon changes in a very short period to conform to the new taxonomy, but it may solve grumbling from this minority. This new taxonomy would presumably update far less often than POWO, which would be nice. But also, this new taxonomy may cause a different group (or even the same one) to be mad about the new taxonomy, too. Also, is there even another global plant taxonomy for us to use? I’m not aware of one.

  4. We permanently freeze taxonomy where it is now (presumably after committing the last several taxonomy changes to align with the current version of POWO, but perhaps not). This option is favored by one particular, very vocal user, but comes with the huge and obvious drawback that iNat taxonomy will be seen by most experts as severely outdated in a few years’ time.

Our current approach to plant taxonomy is, IMO, NOT working and is causing all sorts of problems in a variety of areas. Personally, I feel like the first option is the best, but I feel like we really need to commit to one of them.

12 Likes

Taxonomy isn’t fixed, so 4 is not an option. Even POWO is very slow to roll out changes, and they are well behind the times in so many areas. For my region, following POWO to the letter would be taking approximately 1/5 of the recognized taxonomic entities in the region off the board, which is unacceptable for anyone who does botany down here.

Basically, use POWO as a framework and deviate when necessary. That’s what we’re doing now and it’s complicated but it works for me as a botanist. Plant and plant taxonomy are a VERY messy field with far more species involved than vertebrate taxonomy. That there’s a lot of controversies over various taxa is inevitable. One thing that’s in the works is getting a system in place to start regularly emailing POWO with updates, which should help with some of the deviations.

I believe it’s been stated previously on this forum that we are not going to change from using POWO as a backbone for plant taxonomy.

29 Likes
  1. doesn’t work because there are no other secondary-source global options essentially, and that’s what iNaturalist requires.
  2. cannot work because POWO has some errors (it’s a massive data set, so of course it does). It’s not erroneous enough to throw out entirely, by any means.
7 Likes

I’ve been taking a pretty close look at this, too, in the arena of subspecies. The ones that especially frustrate me are where we aren’t following POWO, there’s no formal deviation in place, and there’s no documentation/flag with any discussion - so there’s no clue why. A lot of those out there. I know a good bit of it is historical processes, etc., so fair enough, but still.

In my opinion, the default for any given species should be POWO (that’s the purported policy!), and if nobody has/can provide a good reason why we have/should deviate, then we ought to be changing whatever isn’t POWO to POWO without much hand-wringing or worry.

4 Likes

One major issue is that a “good reason” to deviate from POWO is completely subjective. That’s actually the main thing tripping us up right now.

2 Likes

As a professional botanist, I think what we have right now is about as good as we are going to get. We have a backbone with POWO and POWO is very responsive when taxonomic issues are brought to their attention. That said, there are thousands of taxonomic issues that may take thousands of years to resolve. There are no clear answers in many situations. We work with what we have and we have deviations to help in some situations. For many situations, we just need to wait around for someone to put in the years of research it takes to resolve most of these types of questions. And, of course, there is a lot of stuff on iNat that needs to be worked out still as far as if it should follow POWO, if POWO should update, or if a deviation is needed. That takes a lot of time too.

I’ve seen the alternative where jobs I’ve worked have required following the USDA Plants Database taxonomy, which is extremely out of date and full of errors. It’s kind of cool in some ways that it doesn’t change but it also means that all the errors and problematic taxonomy gets worse and worse every year. It’s better to slowly update as the changes happen than have a sudden update every 30 years where a huge number of changes all happen at once.

30 Likes

Flag them and get the discussion going. Then there will be a flag and a discussion.

11 Likes

I agree it’s better to update as things happen, as POWO does. But the issue is that this isn’t happening in iNat. Nearly all taxonomic changes that deal with a taxon that has a decently high number of observations just immediately sparks an argument about if it should be followed, leaving nothing to be done, often indefinitely.

4 Likes

Often when there is a flag, there is just an argument in stalemate.

4 Likes

And those are probably the situations where a lot of research still needs to be done because there is no good answer yet. There needs to be a discussion and a decision but the decision often has to be that we’ll just wait and see what happens if someone ever actually researches it. And, of course, not all research is great and further research with additional data could come up with different answers. That is science.

12 Likes

Re 3) World Flora Online exists (it was the original intended successor to The Plant List, but development stalled out for years and POWO became the de facto replacement), but in practice I think it draws a lot of data from POWO.

4 Likes

The couple of problems are:

A. While we’re waiting for that to happen, the list of unmarked deviations from POWO grows longer, and

B. Even once that research happens, there will still be an argument over whether this research is sufficient

2 Likes

They are trying to have as many families as possible supported by TENs. See https://about.worldfloraonline.org/tens. For the rest, they use POWO.

Kew seems to back all of this and tries to keep TENs updates to WFO synced with POWO too. It’s in a weird place right now but I’m guessing/hoping at some point WFO will supercede POWO for taxonomy.

1 Like

The principle problem, I think, isn’t the need for discussion - that’s good - or even the outcome being a stalemate - that’s fine sometimes, too. Even the subjectivity sometimes has its benefits. Personally, I think one of the major problems is that where we have non-formalized deviations, and a flag is raised, the default seems to be inactivity - because nobody chimes in, or because those that do are afraid action will make somebody mad, or because of concern that some expert might eventually come along and tell us why we actually should deviate. A certain amount of caution is good, but the burden needs to be on anybody who wants a deviation to step up and start giving reasons. Otherwise, the default should be action - a reversion to POWO.

I’ve done a bit of this, but I should do some more! I should probably also apply to be a curator, lol. I know some of the issue is that curators are just spread so thin!

I also don’t want to overstate my beefs - haha. We can sort this stuff out! But I sure am glad @raymie brought this up - I’ve been thinking about how to do a forum post on this issue myself, and @raymie really fleshed things out well.

12 Likes

If I remember correctly, the advocate for this approach actually suggested establishment of a committee of varied constituents (taxonomists, ecologists, field biologists) to come to a formal agreement (not freeze taxonomy forever). I assume by vote. That would require some kind of bylaws that would be followed and election of a chair.

3 Likes

Another problem I often experience is that people will make a flag and not tag any relevant curators on the topic. I do my best to make sure I don’t miss any flags for my group but sometimes one will slip by for a while before I notice. A nice solution to this might be a list of the top curators of a taxon and below when making a flag? There might also be a way to get notifications when there’s a new flag on a taxon but I haven’t figured it out yet lol

5 Likes

Another note on not finding flags that show reasons for the taxonomy used. Sometimes the discussion may be on a flag for the genus and cover all the species within the genus. For others the discussion may have been on a taxon that was swapped out and is no longer recognized on iNat. Sometimes it takes some searching around to find these and it is kind of a pain to find the pages for those taxa that are no longer recognized on iNat.

5 Likes

That would be nice. Figuring out how to set up a TEN for one of the groups where POWO (for complex political reasons) does a particularly bad job is on my agenda for the month.

3 Likes

iNat taxonomy is already seen as severely outdated by most experts. It’s been known for many years that dicots were paraphyletic. That taxonomy was outdated when iNaturalist was founded. The lack of monophyletic kingdoms/classes for some organisms was no excuse to use non-monophyletic ones, rather those ranks should have been skipped

iNaturalist probably shouldn’t follow COL. Even now, COL continues to classify dicots as the class Magnoliopsida and put Picozoa in Chromista rather than Plantae. I’d like to see COL dropped as a taxonomic framework and Picozoa moved to its rightful place in Plantae. And I don’t think we should put another framework in COL’s place, as they’re all wrong.

Of course, this is no excuse for letting the inner taxonomy be outdated.

1 Like

For some things yes, but I’ve just as often seen where they quickly and uncritically accept names that are published even if they are nomenclaturally incorrect. Sometimes they’re even unpublished, or ones that were published incorrectly get fixed on POWO but without the correction having been published. I guess it’s better when things get fixed but it kind of drives me crazy when it’s so haphazard, so I’m glad iNat doesn’t follow it strictly.

7 Likes