Recognition of Adiantoideae (Pteridaceae) [PASSED]

[choess]

Author(s) of proposal

Chris Hoess

Name of taxon

Adiantoideae

Rank of taxon

Subfamily

Approximate number of species affected

c. 250

Description of change

The subfamily Adiantoideae, containing the single genus Adiantum, is recognized as distinct from subfamily Vittarioideae, containing the other 11 genera placed in that family in PPG I. The two proposed subfamilies (Adiantoideae and Vittarioideae s.s.) are quite distinct in morphology and phylogenetic evidence now exists to show that both are monophyletic.

Reason for change

The family Pteridaceae, sensu PPG I, contains several morphologically distinctive groups, which have at different times been recognized as families or subfamilies. Among these are two which can be designated the adiantoids and vittarioids.

The former term has been variously circumscribed historically, but is here used for the single genus Adiantum of about 250 species. This genus (and subfamily) possesses an easily recognizable synapomorphy: the margin of fertile leaves is folded over to form a false indusium, and the sori are borne on the folded portion of the leaf lamina. (That is, the opening between false indusium and the body of the lamina is turned away from the margin.) While the degree of leaf division varies widely, species of the genus also tend to have dark, glossy axes (sometimes obscured by indument) and are easily recognized by gross morphology.

The vittarioids, as defined by Crane, consisted of small to medium-sized epiphytic ferns with simple (rarely forked) leaves, generally although not always linear, sporangia following the veins (occasionally connecting between the veins), naked or protected by a groove and/or revolute leaf margin. The morphological coherence of the group is somewhat weakened by the inclusion of Rheopteris, which has pinnate leaves with round, unprotected sori at the margin. Despite its morphological dissimilarity, this genus has historically been grouped with the vittarioids, a decision subsequently validated by molecular phylogeny.

At the time of PPG I, the phylogenetic results available were ambiguous as to whether the adiantoids and vittarioids were sister to one another, or whether Adiantum was nested within the vittarioids. Accordingly, subfamily Vittarioideae in that classification was circumscribed to include both groups to ensure its monophyly. More recent results (e.g., Schuettpelz et al. 2016) have returned these two groups as monophyletic and sister to one another.

The relative morphological coherence of the adiantoids and vittarioids when considered separately, and their incoherence when joined into a single group, makes it desirable to recognize both taxonomically as Adiantoideae and Vittarioideae, respectively.

Furthermore, several of the vittarioid genera are fully separable only on microscopic characteristics (morphology of spores and soral paraphyses), so photographic and sterile material may not be identifiable to that level. In organizing imperfect specimens and data, it is convenient to have a formal taxonomic group that corresponds to this gross morphology but is not intermingled with the distinctive Adiantum. This separation also allows the subfamilies to be readily separated in keys.

While it might be argued that is undesirable to erect a subfamily for a single genus, that genus is highly recognizable, speciose, and also diverse in morphology. Separating Adiantoideae and Vittarioideae provides a taxonomic system that is overall more useful and coherent than that which had to be adapted in PPG I.

• Crane, Edmund H. (1997) A Revised Circumscription of the Genera of the Fern Family Vittariaceae. Syst. Bot. 22(3): 509-517. https://doi.org/10.2307/2419824
• Huiet, Layne et al. (2018) A worldwide phylogeny of Adiantum (Pteridaceae) reveals remarkable convergent evolution in leaf blade architecture. Taxon 67(3): 488-502. https://doi.org/10.12705/673.3
• Schuettpelz, Eric et al. (2007) A molecular phylogeny of the fern family Pteridaceae: Assessing overall relationships and the affinities of previously unsampled genera. Molec. Phylogen. Evol. 44: 1172-1185. https://doi.org/10.1016/j.ympev.2007.04.011
• Schuettpelz, Eric et al. (2016) A revised generic classification of vittarioid ferns (Pteridaceae) based on molecular, micromorphological, and geographic data. Taxon 65(4): 708-722. https://doi.org/10.12705/654.2

Reference(s) for publication of the name

Adiantoideae Horvat Acta Bot. Inst. Bot. Univ. Zagreb. 2: 114 (1927), as "Adiantinae".

"Die dritte Unterfamilie stellen die Adiantinae vor."

Horvat adiantoideae.pdf

List the numbers of any related issues

No response

Code of Conduct

• [X] I agree to follow the PPG Code of Conduct

[choess]

I am trusting a bit in IPNI that Horvat established the subfamily here, but he does seem to have employed the correct rank-denoting term (quoted).

In the preceding discussion, there was some consideration of breaking up Pteridaceae, a treatment with some historical precedent. I am not definitely opposed to that idea, but I didn't feel ready to write that proposal. I have put this forth in the hopes of establishing a separate Adiantoideae; it should not be a bar to further proposals if anyone so desires.

[SchneiderHarald]

The treatment in PPGI reflected some of the uncertainties concerning the monophyly of Adiantum which are now more or less resolved. Adiantum forms a monophylum which is sister to the remaining vittarioideae which are corresponding more or less to the former family Vittariaceae. Thus, the proposal makes sense on the first sight but I am not in full support here. The split into two subfamilies ignored the fact that Vittarioideae according to PPGI form a very robust clade. Instead, we should consider to utilize the full breath of Linnaean hierarchies available. Thus, I suggest to reject this proposal and replace it with a proposal introducing two tribes. This approach is much more suitable to represent the phylogeny recovered.

[crothfels]

Personally, I'm in favor of this proposal. For example, it would be very helpful to have a taxon available for the vittarioids s.str. (ie, excluding Adiantum) when identifying observations in iNaturalist. That fact that Adiantum + vittarioids s.str. is a well supported clade doesn't bear on the appropriate rank in my opinion (a subfamily can be sister to a clade of subfamilies, just as a family can be sister to clade of families. Cystopteridaceae being sister to the rest of the eupolypods II doesn't mean that we should recognize that clade [the rest of EuII] as a family comprising a bunch of subfamilies...), and I don't see a benefit of introducing rarely used ranks when subfamily works just as well, or better.

[choess]

I like the general idea of making more use of intermediate ranks like tribes to improve classification (the Microsoroideae come immediately to mind). That said, given the limited number of ranks available, I think it makes sense to reserve them for clades with well-defined and useful synapomorphies; otherwise, it seems unlikely that there will be much use for a name attached to the clade. I'm happy to be corrected, but I don't know of any morphological or molecular biological synapomorphies that strongly unite Adiantum and the vittarioids s.s.

This proposal was inspired in part by iNaturalist. Since this is effectively a collection of photo vouchers, it's not uncommon to see photos that are assignable to vittarioids s.s. but can't be systematically keyed further. (They might be identifiable from locality, habitat, and other context by one familiar with them.) That probably extends to sterile herbarium material as well, and I think it would be nice to have a label for determining these that doesn't commingle them with the easily-distinguishable Adiantum.

[msundue]

There are easily observed characters I know of that unite Adiantum and the Vittarioids s.s. First, they have reddish young leaves. This is homoplastic with other ferns (e.g. Blechnaceae and others) but still useful. They also all bear elongate silica bodies (called venuloid idioblasts in older literature). Again, there is homoplasy depending on how you parse the character, but the silica bodies of Adiantum and Vittaroids s.s. are particularly similar morphologically. Take a look at fig. 1 in Sundue, M., 2009. Silica bodies and their systematic implications in Pteridaceae (Pteridophyta). Botanical Journal of the Linnean Society, 161(4), pp.422-435.

Sundue 2009.pdf

[gonggashan88]

Anyone prefers these taxa as families over subfamilies? These subfamilies have a long history of being recognized as families. All relevant familial names exist. Taxonomic ranks are arbitrary. Well, the currently defined Pteridaceae s.l. are too large, which is not very friendly for communication, monographic work, and further analysis.

Unlike other large families, such as Dryopteridaceae and Polypodiaceae, in which (some) subfamilies are not clearly defined morphologically, nearly all current subfamilies in Pteridaceae s.l. are quite distinct morphologically.

We have only 50+ families for pteridophytes vs. bryologists' 120+ families for similar number of species of mosses.

Adiantaceae Cryptogrammaceae Hemionitidaceae Parkeriaceae Pteridaceae s.s. Vittariaceae

[apprenti-sorcier]

Taxonomic ranks are arbitrary. Well, the currently defined Pteridaceae s.l. are too large, which is not very friendly for communication, monographic work, and further analysis.

Unlike other large families, such as Dryopteridaceae and Polypodiaceae, in which (some) subfamilies are not clearly defined morphologically, nearly all current subfamilies in Pteridaceae s.l. are quite distinct morphologically.

We have only 50+ families for pteridophytes vs. bryologists' 120+ families for similar number of species of mosses.

I don’t think this was a good example.

Or, then let’s also take the example of the Angiosperm Phylogeny Group: at the recent International Botanical Congress, the people who have led APG for over 25 years explained that they have always tried to be as inclusive as possible with families, to assemble rather than disassemble. Parsimony should be kept in mind.

I mean the principle of Ockham’s razor, which states that while two or more equally robust options co‐exist, the simplest is the best. The simplest is the less disruptive one to the established uses; in practice it’s the one needing the least new names or combinations.

To this day there are people who prefer to recognise Fumariaceae or Chenopodiaceae and others as families within APG rather than subfamilies because they are undoubtedly monophyletic and morphologically distinctive.

Nevertheless, the success of APG in the botanical community shows us its principles are inspiring. Flora authors and others non‐pteridologists (as many participants expressed concerns at the recent PPG TEN meeting in Madrid) still wonder why APG simplified the family taxonomy while PPG Ⅰ splits so much relatively to it.

[gonggashan88]

@apprenti-sorcier, thanks for the comments. Well, your examples of Fumariaceae and Chenopodiaceae are hardly relevant, aren't they? In both cases, recognition of Fumariaceae and Chenopodiaceae would make Papaveraceae and Amaranthaceae, respectively, paraphyletic unless a third family, Pteridophylloideae and Polycnemaceae, respectively, would be recognized but they each contain 1 or a few species only. This is in no way comparable to the case of six subfamilies/families in Pteridales that each have a long history of recognition and each are monophyletic and species-rich.

We can learn from the APG classifications, but ferns are not flowering plants. Bryologists continue to recognize 120+ families in mosses for the similar number of species in ferns, no matter how APG people classify flowering plants!

Again, taxonomic ranks are arbitrary. Traditional dominant fern classifications just happened to recognize very limited number of families, from Hooker (18xx), Kramer (1990), Smith et al. (2006), to APGI (2016). Whether this is the best idea is questionable in terms of informativeness and convenience in promoting communication and further studies.

[algrusz]

I appreciate this important discussion; here are my two cents: Although I agree that a taxonomic handle for Adiantum and vittarioids (s.s.) would be helpful, I also agree that the strong molecular and, albeit subtle, morphological synapomorphies uniting the Vittarioideae deserve to be recognized. For this reason, I am in favor Harald's suggestion to retain the Vittarioideae "as is" and to differentiate among the sister clades at the level of tribe.

[joelnitta]

This proposal was voted on during PPG Ballot 13 (voting period September 2024). A total of 82 votes were cast. There were 65 'Yes' votes (79.3%) and 17 'No' votes (20.7%). The proposal passes.

[apprenti-sorcier]

your examples of Fumariaceae and Chenopodiaceae are hardly relevant, aren't they? In both cases, recognition of Fumariaceae and Chenopodiaceae would make Papaveraceae and Amaranthaceae, respectively, paraphyletic unless a third family, Pteridophylloideae and Polycnemaceae, respectively, would be recognized but they each contain 1 or a few species only

@gonggashan88, sorry for the digression; in the first case the consensus summarised by the Caryophyllales Taxonomic Expert Network is that until further science, only Chenopodiaceae can now be confidently separated from the Amaranthaceae, which they do (https://caryophyllales.org/cdm_dataportal/taxon/62dec049-7461-4942-8d87-4b0506c01f45) In the other case, Papaveraceae sensu APG can be also divided into Papaveraceae s.s., Fumariaceae as traditionally viewed and two small families Pteridophyllaceae, Hypecoaceae (https://www.researchgate.net/publication/367329314_A_complete_genus-level_phylogeny_reveals_the_Cretaceous_biogeographic_diversification_of_the_poppy_family/figures); they are monophyletic.

Now let’s go from Bryophytes and Angiosperms back to the ferns.