Taxonomic novelties in grammitid ferns (Polypodiaceae) from the Neotropics and Madagascar supported by molecular data

Abstract

Based on cpDNA data, we provide the phylogenetic position for 18 species of Neotropical grammitid ferns (Polypodiaceae) that were not previously included in a molecular phylogeny. These species were resolved in Alansmia, Ceradenia, Enterosora, Grammitis, Lellingeria, Lomaphlebia, Melpomene, Moranopteris, Stenogrammitis, and Terpsichore. Our results indicate that Enterosora is polyphyletic and in need of generic recircumscription. We maintain the identity of Enterosora, based on the position of E. campbellii subsp. spongiosa, a variety of the type species. This finding allowed us to conclude that the E. parietina clade should be excluded from Enterosora. To accommodate this clade we describe a new genus, Parrisia. It is related to Adenophorus and a clade that includes Cochlidium, Grammitis s.s., and Lomaphlebia. We further found that Zygophlebia is paraphyletic with respect to Enterosora, represented by two separate clades, Zygophlebia s. s. that includes Z. cornuta, the type species, Z. sectifrons and Z. matthewsii, and a second clade that includes another seven species of Zygophlebia as well as E. barbatula. As a separate clade, this Zygophlebia species could be maintained. However, we prefer to sink both Zygophlebia clades into Enterosora. We propose ten novel combinations for Enterosora, by moving Zygophlebia with a confirmed phylogenetic position to Enterosora. It is possible that the remaining Zygophlebia species should also be treated as Enterosora, but further research is necessary.

A global phylogeny of grammitid ferns (Polypodiaceae) and its systematic implications

“…and Grammitis s.l. has been well documented in earlier molecular works (Ranker & al., 2004; Perrie & Parris, 2012; Sundue & al., 2014; Bauret & al., 2017; Shalisko & al., 2019), but it had not been clear to which clades the two generic names should be applied because neither of the type species had been sampled in previous studies. There are specimens of the two type species in some herbaria, but they are too old or not well preserved for Sanger sequencing.…”

“…The type genus of Grammitidoideae, Grammitis s.l., has long been known to be polyphyletic (Ranker & al., 2004; Schuettpelz & Pryer, 2007; Sundue & al., 2010, 2014; Bauret & al., 2017; Shalisko & al., 2019). Bauret & al.…”

“…At the generic level, molecular studies have sampled 31 out of the 36 genera recognized and the only 5 genera that have not yet been sampled are Archigrammitis and Luisma (Bauret & al., 2017; Shalisko & al., 2019) plus the three newly described genera, Calligrammitis , Devolia , and Glabrigrammitis (Yang & al., 2023). Molecular data have also supported the monophyly of 27 genera in Grammitidoideae, whereas 3 genera, Ctenopterella , Grammitis , and Oreogrammitis sensu Parris & Sundue (2020), have not been supported as monophyletic, with Ctenopterella and Grammitis being highly polyphyletic (Ranker & al., 2004; Sundue & al., 2014; Bauret & al., 2017; Shalisko & al., 2019). Only one species of Acrosorus Copel.…”

“…(2022), who included two species and supported the monophyly of the genus. Previous studies included 27 type species of accepted genera (Ranker & al., 2004; Perrie & Parris, 2012; Sundue & al., 2014; Bauret & al., 2017; Shalisko & al., 2019), whereas the remaining 9 type species had not yet been included in earlier molecular works (Appendix 1). Those nine type species included those of the highly polyphyletic Ctenopterella and Grammitis , preventing the accurate delimitation of these species‐rich genera.…”

“…At the species level, earlier studies were based on relatively large taxon sampling by including 202, 266, and 396 accessions representing 190, 221, and 224 species by , Shalisko & al. (2019), and , respectively.…”