APG V changes in family classification

Curators
1

The Angiosperm Phyllogeny Group has released an update from the APGIV classification currently used on iNat: https://onlinelibrary.wiley.com/doi/10.1111/jse.70096

Per the abstract, here are the large changes:

At the family level, expanded circumscriptions include Tecophilaeaceae (including Ixioliriaceae, Asparagales), Frankeniaceae (including Tamaricaceae, Caryophyllales), Phytolaccaceae (including Agdestidaceae and Sarcobataceae, Caryophyllales), Helwingiaceae (including Phyllonomaceae, Aquifoliales), Gesneriaceae (including Calceolariaceae and Peltantheraceae, Lamiales), Pentaphylacaceae (including Sladeniaceae, Ericales), and Orobanchaceae (including Mazaceae, Paulowniaceae, Phrymaceae, and Wightiaceae, Lamiales). In Santalales, the limits of Santalaceae and Olacaceae are revised, and Erythropalaceae and Strombosiaceae are recognized as distinct from Olacaceae

For curation at these high levels above genus, does iNat still follow POWO? When should we start implementing and would it be best to have all the discussion here on the forum or should a flag be created for every change?

For most of these changes, I think we should try and implement by regrafting to maintain taxa whenever possible and change ranks rather than conducting merges e.g. Mazaceae becomes subfamily Mazoideae under Orobanchaceae.

EDIT: I’ve tallied all the changes from APGV which affect iNat, please let me know if I’m missing anything. These total 31 taxa which need re-grafted or merged affecting 457,000 total observations

Rank Taxon Change description link affected taxon Num. affected iNat Observations
Family Peraceae Merged into Euphorbiaceae https://www.inaturalist.org/taxa/71594-Peraceae 8100
Family Tamaricaceae Merged into Frankeniaceae https://www.inaturalist.org/taxa/51304-Tamaricaceae 56000
Family Calceolariaceae Merged into Gesneriaceae https://www.inaturalist.org/taxa/71463-Calceolariaceae 22000
Family Phyllonomaceae Merged into Helwingiaceae https://www.inaturalist.org/taxa/71601-Phyllonomaceae 100
Genus Cassinopsis Moved into Metteniusaceae https://www.inaturalist.org/taxa/576230-Cassinopsis 400
Family Mazaceae Merged into Orobanchaceae https://www.inaturalist.org/taxa/571193-Mazaceae 25000
Family Paulowniaceae Merged into Orobanchaceae https://www.inaturalist.org/taxa/71586-Paulowniaceae 42000
Family Phrymaceae Merged into Orobanchaceae https://www.inaturalist.org/taxa/64552-Phrymaceae 280000
Family Sladeniaceae Merged into Pentaphyllaceae https://www.inaturalist.org/taxa/71624-Sladeniaceae 10
Family Sarcobataceae Merged into Phytolaccaceae https://www.inaturalist.org/taxa/71421-Sarcobataceae 7600
Family Apodanthaceae Merged into Rafflesiaceae https://www.inaturalist.org/taxa/71405-Apodanthaceae 400
Family Bataceae Merged into Salvadoraceae https://www.inaturalist.org/taxa/70041-Bataceae 8000
Family Ixioliriaceae Merged into Tecophilaeaceae https://www.inaturalist.org/taxa/71546-Ixioliriaceae 1500
Genus Brachynema Move to new family Erythropalaceae https://www.inaturalist.org/taxa/576353-Brachynema 0
Genus Erythropalum Move to new family Erythropalaceae https://www.inaturalist.org/taxa/427623-Erythropalum 60
Genus Heisteria Move to new family Erythropalaceae https://www.inaturalist.org/taxa/186461-Heisteria 1200
Genus Maburea Move to new family Erythropalaceae https://www.inaturalist.org/taxa/1069762-Maburea 0
Genus Diogoa Move to new family Strombosiaceae https://www.inaturalist.org/taxa/1070041-Diogoa 0
Genus Engomegoma Move to new family Strombosiaceae https://www.inaturalist.org/taxa/1070068-Engomegoma 0
Genus Scorodocarpus Move to new family Strombosiaceae https://www.inaturalist.org/taxa/348274-Scorodocarpus 50
Genus Strombosia Move to new family Strombosiaceae https://www.inaturalist.org/taxa/133181-Strombosia 150
Genus Strombosiopsis Move to new family Strombosiaceae https://www.inaturalist.org/taxa/1069868-Strombosiopsis 10
Genus Tetrastylidium Move to new family Strombosiaceae https://www.inaturalist.org/taxa/1069777-Tetrastylidium 0
Family Cardiopteridaceae Move to new order Cardiopteridales https://www.inaturalist.org/taxa/71468-Cardiopteridaceae 1800
Family Stemonuraceae Move to new order Cardiopteridales https://www.inaturalist.org/taxa/71630-Stemonuraceae 300
Family Oncothecaceae Move to new order Oncothecales https://www.inaturalist.org/taxa/71582-Oncothecaceae 20
Genus Borthwickia Moved to Resedaceae https://www.inaturalist.org/taxa/1069954-Borthwickia 0
Family Forchhammeria Moved to Resedaceae https://www.inaturalist.org/taxa/209917-Forchhammeria 1000
Family Columelliaceae No longer in Bruniales, regraft to Magnoliopsida to account for its unplaced status https://www.inaturalist.org/taxa/71479-Columelliaceae 1700
Family Huaceae No longer in Oxalidales, regraft to Magnoliopsida to account for its unplaced status https://www.inaturalist.org/taxa/71536-Huaceae 10
2

You may notice that the major authors of APG V are also those that led GLOVAP:

https://forum.inaturalist.org/t/how-do-we-deal-with-the-thousands-of-highly-contested-glovap-new-combinations/10277

In short, no, I would not implement them.

3

And also largely the same authors as APG IV which we currently follow.

I do not think that this work should be dismissed because of any affiliations to other work but should be judged on its own merit.

4

Have you checked with staff to see if these are even possible? Obviously splitting magnoliids out of dicots is still out (if anyone doesn’t know why, see prior discussions about the limits of iNat’s processing capabilities), but lumping Phrymaceae into Orobanchaceae seems potentially very problematic based on what staff have said about prior large changes. The double-crested cormorant change apparently nearly broke the site, and that was 90k observations. The genus Erythranthe alone has 133k.

5

Good point, I have not. @loarie what’s your take on APGV and how we should roll it out on iNat?

6

iNat follows POWO as its external reference for plants. They only go up to Family. As mentioned above family we likely get into infrastructure constraints where iNat can’t currently handle major changes at the base of the tree. But the family taxonomy should reflect the reference. Is POWO planning on incorporating APGV changes?

Separately, I’ve heard that WFO may be a more suitable external reference for iNat than POWO. Would there be more buy in from botanists on iNat if we were following WFO rather than POWO?

7

Do you have an estimate for exactly how many affected observations constitute a “major change”? If a change did exceed that number would it mean it would have to be implemented by admins or it is not implementable at all without infrastructure upgrades?

In the context of APGV, I expect that POWO and WFO online will implement it the same way that APGIV has largely been accepted by most taxonomic authorities, but I have not heard directly.

8

@tiwane will have a better sense for me of where to find that int he support documentation. But there’s current curator restrictions on which taxa can be swapped and which ancestries can be changed based on the number of downstream observations meant to avoid curators triggering such major changes.

Unfortunately, staff aren’t able to make these major changes at the moment either. The changes that would need to be made are improvements to the backend infrastructure. Because currently, when the taxonomy changes and there are lots of downstream observations the task of updating all of those observations isn’t scaling well.

9

It looks like the largest change is Phyrmaceae which currently has 281,000 observations. The other large family level changes are closer to the 50,000 observations range or smaller.

EDIT: I’ve since tallied the exact changes, there are 31 taxa which need re-grafted or merged tallying 457,000 total observations

10

A couple years ago a series of flags were opened to discuss interest in transitioning to WFO (Angiosperms, Gymnosperms, Pteridophytes, Bryophytes).

My impression is that as of early 2025 there was interest in using WFO as the standard for pteridophytes and bryophytes since POWO either doesn’t apply to them or curates them poorly. However for angiosperms and gymnosperms the preference was to stay with POWO because POWO stays more up-to-date and WFO hadn’t caught up yet. From the flag comments it sounds like POWO and WFO might eventually hope to merge together, but that would still be years from now.

The pteridophyte and bryophyte flags have been open since then without resolution, I assume waiting for go-ahead from staff although nobody has actually explicitly asked for it yet.

11

Just taking a quick look at WFO for the taxa I worked on for my PhD, while POWO accepts my treatment, WFO has all of the new taxa and name changes I made as “unplaced” and the taxa/ranks they accept are a mishmash of multiple treatments. “Unplaced” taxa seem like a problem. Like POWO, WFO don’t really give any justification for what they choose to do. Seems like WFO has potential but may still be well behind POWO (at least for some taxa). It’s probably all a big mess one way or another though and that is why we have deviations.

12

I’m the one who wrote that whinge-fest, but in all honesty, I don’t consider APGV to be all that controversial, or a similar issue. IMO GLOVAP was a self-published work that exploited uncertainty to re-combine vast swathes of species to create combinations that may well be short-lived.

In my opinion, the larger family classifications are largely well justified by data, and higher ranks have always been more abstract anyway.

I have no problem with this.

13

Given the strong overlap with Kew personnel, I think that is a safe assumption.

I was a respondent to the APG V survey that was sent out prior to publication. My concerns centered mainly around Boraginaceae s.latiss. (=Boraginales) as did those of many other respondents, but also around the new merger of Orobanchaceae + Phrymaceae et al. As best I can tell, the latter renders the expanded Orobanchaceae morphologically unrecognizable and undiagnosable (even more so than the already heterogeneous Boraginaceae).

These family lumpings are especially concerning because

The APG stability principle states that if a currently accepted APG family or order has been shown to be monophyletic, it will not be subsequently split.

(see the publication). This principle creates a one-way “ratchet,” leading only to ever more expansive, unrecognizable, and meaningless “families” in future iterations of APG as additional molecular and phylogenetic data accumulate.

Clearly APG on some level still recognize the value of smaller, more recognizable families. Otherwise, if “stability” were the only imperative, we could just call all angiosperms Magnoliaceae and be done with it.

For these reasons I would counsel against haste in adopting the latest APG even after POWO has incorporated it. Even though APG state that they will never undo their family lumpings, it might not hurt to stay one iteration behind the most recent while trends of acceptance or rejection in the larger systematics community can be gauged. Unless, I suppose, having a lot of “alternate positions” in our Taxon Framework Relationships puts undue burden on iNat infrastructure.

At the least, I would prefer to see each proposed change discussed on its own taxon flag before any decision to adopt it, instead of implementing all APG V changes in a single action.

I take heart (maybe perversely) in knowing that

I’m not-so-secretly hoping that this is a bridge too far for current iNat infrastructure, and will at least delay the merger into Orobanchaceae for a few years.

14

As someone who’s had to deal with the fallout from GLOVAP in fern taxonomy for years, I don’t think that this should be a bar to APG V implementation. My concerns largely track those expressed by @jdmore . The Boraginaceae issue has been a particular sore point for years; the expansion of Orobanchaceae definitely changes the character of that family, and we shall have to see if that choice sticks, but it sounds like the site can’t handle the change at present, anyway.

15

I was part of some conversations with staff over a year ago (I think?) which seemed to come to a positive resolution on switching pteridophytes to WFO, but it then fizzled out.

I’ve since become part of the PPG II coordinating committee. PPG is the taxonomic expert network recognized by WFO for ferns and lycophytes, and the new PPG II classification (to genus level) is in proof about to be published. We’ve assembled an array of curators to maintain things down to species level in WFO, so I think I can say we are on top of things.

I’ve been waiting for the manuscript to actually hit print and have a breath of free time, but lobbying staff again to switch ferns & lycophytes to WFO is definitely on my to-do list.

16

I do not get the need for the family lumping. Can’t the clade be a superfamily?

17

Looking at the APG paper itself, it has supplementary data which includes a broad assessment of opinions from a survey. Only 61.8% of 228 respondents agreed with >80% of the classification, though 16.7% had no opinion. Either way, that’s not very high support.

image
image631×290 26.9 KB

They follow this with this baffling statement related to the question “Name up to 5 groups (either families or orders) with which you strongly disagree and indicate why.”;

Note that 20 of the 124 respondents to this question stated no disagreements with APG V; thus, 120 of the 124 respondents to the survey had no disagreements with the classification, and the disagreements noted below are based on 104 responses.

If their level of review of the supplemental data is as good as their review of APG V, I can understand why there might be some disagreements. 84% of respondents strongly disagreed with at least one family or order from APG V.

18

I can understand that association [of the authors] doesn’t automatically negate their decisions, but also, I still don’t agree…

For example, I don’t understand the need to lump Phyrmaceae into Orobranchaceae simply because they form a clade together, even though they’re independent separable. As vladimir_fuka said earlier, why can’t they just be in a epifamily together? Anyways-- it looks like iNat’s systems wouldn’t be able to handle such a move.

19

I’m just an undergrad and trying to understand what’s going on. When the paper indicates the clade containing Asparagales and Ixioliriaceae is being circumscribed as a new single family, that means all of the Asparagales, right? As in, orchids and yuccas and irises and so on? That entire order, plus a small family, as a single mega-family? The Asparagaceae alone was already really broad under APG III; as a group it is morphologically meaningless if you lump it in with the orchids

Edit: okay, I went and read the paper and got my answer. Good reminder to always do your due diligence :)

20

I know other people have already made similar points, but I thought I’d throw in that I disagree with APG’s philosophy and find it totally counter to how angiosperm families are applied practically, especially on iNat. There’s no point to implementing a lot of these changes here because the groups will still remain as infrafamilial taxa (e.g. our Asparagaceae). All this fuss over arbitrary ranks to have people continue to use the same groups they were before, just with new pigeonholed names and ranks. It’s a lot of work for no benefit to us.

@kevinfaccenda just a note, the numbers in your table appear to only be for Research Grade observations. There are 10,000 more affected observations for Phrymaceae alone.

@rowanbynight you are meant to read that as “we are expanding Tecophilaeaceae to include Ixioliriaceae, which was formerly recognized as a family in Asparagales.” Further down the paper there is a table with their complete recommended treatment that shows APGV accepts 13 families in Asparagales. Also, as a side note, APG’s Asparagaceae was already pretty meaningless as a family anyway.