Combining the small South American genus Androtrichum into Cyperus (Cyperaceae)

1Programa de Pós-Graduação em Biologia de Fungos, Algas e Plantas, Universidade Federal de Santa Catarina, Departamento de Botânica, Campus Universitário, Trindade, Florianópolis, SC, 88040-900, Brazil 2Universidade Federal de Santa Catarina, Departamento de Botânica, Campus Universitário, Trindade, Florianópolis, SC, 88040-900, Brazil 3Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3AE, UK 4Ghent University, Department of Biology, Systematic and Evolutionary Botany Lab, K.L. Ledeganckstraat 35, 9000 Gent, Belgium *Corresponding author: silvapeluciana@gmail.com REGULAR PAPER

Androtrichum was described based on the species Abildgaardia polycephalum Brongn. [= Androtrichum trigynum (Spreng.) H. Pfeiff.]. Although the habit of this species differs from that of Abildgaardia Vahl, the structure of its spikelets at the time of flowering is somewhat similar which led Brongniart (1833) to publish Androtrichum first as a subgenus of Abildgaardia: "Abildgaardia subg. Androtrichum Brongn.: Filamentis staminum in fructu elongatis, longe exsertis (an Genus proprium?)". Since then, more specimens of this species were collected including specimens in fruit. Those specimens present a particular character, i.e., the considerable elongation of the stamen filaments. The filaments, which because of their sheer number, their whiteness, and their length have the appearance of hairs (figs 1D, 2E), arise from the glumes and they remain on the fruits when these fall from the plant when mature. This character, combined with a rhizomatous habit differing from that of Abildgaardia, was deemed sufficient to accept Androtrichum as a separate genus (Brongniart 1834).
However, other authors placed this species (A. trigynum) in no less than six different genera besides Androtrichum, i.e., Abildgaardia, Comostemum Nees, Cyperus, Eriophorum L., Scirpus Tourn. ex L., and Trichophorum Pers. From this list, it can be assumed that this species combines several more or less conspicuous characters of all these genera. Pfeiffer (1940) proposed the inclusion of a second species in Androtrichum (Scirpus giganteus Kunth [≡ A. giganteum (Kunth) H.Pfeiff.)]), which also presents elongation of the stamen filaments (figs 1B, 2B). Govaerts et al. (2007Govaerts et al. ( , 2020 accepted the inclusion of the second species in Androtrichum, although Goetghebeur (1998) did not consider its inclusion to be convincingly demonstrated.
In the molecular phylogenetic study of Muasya et al. (2009a), Androtrichum trigynum was retrieved as sister to the Cyperus Clade, while A. giganteum was positioned in a polytomy within the Cyperus Clade. Using the same accessions sequenced for the chloroplast gene rbcL only as part of a study including a more extensive sampling of Cyperaceae, Semmouri et al. (2019) found that both species are clearly placed within the C 3 Cyperus Grade of the Cyperus Clade. Van der Veken (1965) and Goetghebeur (1986) both studied the embryo of A. trigynum and noted that it corresponds to the Cyperus-type (Semmouri et al. 2019). However, the C 3 photosynthetic pathway has only been confirmed for A. trigynum (Larridon et al. 2011a). To date, the Androtrichum species have not yet been formally combined in Cyperus, which remains paraphyletic, although a monophyletic genus Cyperus is supported based on molecular data. In this study, we investigate if the morphology and anatomy of Androtrichum species provide additional evidence for their placement in the genus Cyperus to delimit it as a monophyletic group. In addition, we provide descriptions and distribution data for the species.

MATERIAL AND METHODS
The morphological study was conducted using specimens from the herbaria FLOR, FURB, GENT, ICN, and K (acronyms according to Thiers continuously updated). Images of spikelets and fruits were captured using a digital camera attached to a stereoscope microscope (Leica EZ4 D).
The anatomical study was carried out on samples of Androtrichum trigynum (culm) and A. giganteum (leaves and culm), collected in the state of Santa Catarina, Brazil. The samples were fixed in formaldehyde-acetic acid-ethanol 70% (FAA 70) (Johansen 1940) for 24 hours and were later stored in ethanol 70%. Mature middle regions of leaves and culm were sectioned free-hand with a razor blade and stained with 1% Safranin and 1% Astra Blue (Bukatsch 1972). The anatomical characters were analysed, and images were captured using a digital camera (Leica DFC 295) attached to a microscope (Leica DM 2500).
The taxonomic changes and typifications of names were performed according to the International Code of Nomenclature for algae, fungi, and plants (Turland et al. 2018).

RESULTS
The morphology of the Androtrichum species agrees with their placement in the C 3 Cyperus Grade of the Cyperus Clade. Androtrichum species share a combination of morphological characteristics with the C 3 Cyperus species such as anthelate inflorescences, bisexual flowers, trigonous achenes ( 3C). In A. trigynum, the culm has a uniseriate epidermis with the outer periclinal wall conspicuously thickened and covered by a cuticle (fig. 3D). Below the epidermis, several strongly thickened fibres strands are   3D). In the epidermis, stomata are observed with large substomatal chambers in the direction of the chlorophyll parenchyma ( fig. 3D). The vascular system consists of collateral vascular bundles, which tend to increase in size towards the centre of the organ (fig. 3D). All vascular bundles are surrounded by a pluricellular sheath of thickened cells (fig. 3D). The pith region has ground parenchyma with small cells and few intercellular spaces ( fig. 3D).
Based on evidence from morphology and anatomy that supported the genus Androtrichum as part of the C 3 Cyperus Grade of the Cyperus Clade, we combine the two species of the genus Androtrichum into Cyperus. For A. trigynum, a combination in Cyperus is already available: C. trigynus Spreng. For A. giganteum, a new name in Cyperus is published: C. byssaceus Pereira-Silva nom. nov. A taxonomic treatment is presented below with the relevant typifications, detailed morphological descriptions, and distribution data.

DISCUSSION
Due to its superficial resemblance to several other genera in Cyperaceae, the affinities of Androtrichum have long remained unclear. Since Van der Veken's (1965) embryo morphology study, we know that A. trigynum has an embryo of the Cyperus-type. In his revision of the Cyperaceae genera, Goetghebeur (1986) indicated that the inflorescence of A. trigynum agrees with the general inflorescence blueprint of Cyperus. Goetghebeur (1986) also remarked on the striking similarities between Androtrichum and the genus Scirpoides. Thus, although the presence of strongly elongated filaments after anthesis is not observed in any other species of Cypereae, there was little doubt about the position of Androtrichum near this tribe (Goetghebeur 1986). Muasya et al. (2009a) confirmed the close relationship of A. giganteum and A. trigynum with Cyperus, but it remained unclear whether (1) the genus Androtrichum was monophyletic, and (2) whether it was sister to or nested in Cyperus.
Recent molecular phylogenetic hypotheses confirmed the placement of both species of Androtrichum in the C 3 Cyperus Grade and that it is a polyphyletic genus (Semmouri et al. 2019). In addition, the morphological characteristics of Androtrichum species agree with the general morphology of the C 3 Cyperus species. Our anatomical study confirms that A. trigynum and A. giganteum do not possess Kranz anatomy, which provides additional arguments for its placement among the C 3 Cyperus lineages. Furthermore, the two Androtrichum species have common anatomical features found in most Cyperus species, such as V-shaped leaves, large and conspicuous air cavities in the leaf blades, numerous vascular bundles embedded in the chlorenchyma or in its internal limit, and sclerenchyma as crescentiform caps of fibres at xylem poles and usually as hypodermis in the culm (e.g., Metcalfe 1971;Denton 1983;Hefler & Longhi-Wagner 2010). Based on that, we formally synonymize A. giganteum and A. trigynus into Cyperus. With the taxonomic changes proposed here, Cyperus is delimited as a monophyletic group and the only genus of the Cyperus Clade of tribe Cypereae.

TAXONOMIC TREATMENT
Cyperus L. (Linnaeus 1753: 44). Abildgaardia subg. Androtrichum Brongn. (Brongniart 1833: 176). -Androtrichum (Brongn.) Brongn. (Brongniart 1834: 177 35-90(-150) cm × 20-35 mm, leaf-like, scabrous margins. Inflorescence a terminal lax anthela with several capitate partial inflorescences up to third order; primary branches 10-15, lowermost branches 8-25 cm long. Spikelets 4-8 × 2-2.5 mm, with 8-15 glumes; rachilla persistent. Glumes (2.6-)2.9-4.8 mm long, spirally arranged, deciduous, slightly mucronate, mucro up to 0.7 mm long, reddish stramineous to reddish light brown. Stamens 3, filaments reddish brown, strongly elongating after flowering; anthers 1-2 mm long. Style deeply trifid; style base slightly thickened, persistent. Achenes 0.9-1.2 × 0.3-0.5 mm, elliptical to narrow obovate, trigonous, brown, surface puncticulate. Distribution and ecology -Cyperus byssaceus occurs in Argentina, southern Brazil, Paraguay, and Uruguay (Govaerts et al. 2020). It is most often found in large populations in silted marshes and along the banks of creeks and small streams. Notes -Based on Article 53 of the ICN (Turland et al. 2018), the new name Cyperus byssaceus is proposed for A. giganteum, since the name Cyperus giganteus Vahl (Vahl 1805) is an earlier validly published name of another species. The epithet "byssaceus" was given due to the filamentous appearance of the inflorescence caused by the elongation of the stamen filaments. Additional material examined -Argentina: Buenos Aires: Campana, 27 Nov. 1938 (Govaerts et al. 2020). A halophytic plant, which is characteristic of the vegetation of coastal dunes. Cyperus preslii was described based on a specimen from Peru collected by Haenke. In the PRC herbarium, there is a Haenke specimen, however, the location of the collection is not mentioned, which makes it impossible to confirm whether this specimen is from Peru, since no other record of this species has been made for this country (see also Pfeiffer 1937). Notes -Both the protologue of Cyperus trigynus and of Eriophorum montevidense state that the description was based on a specimen from Montevideo. The protologue of E. montevidense further details it to be a Sellow specimen. The original material was studied at the B herbarium, with duplicates (labelled as such) sent to K and P. The original material was likely destroyed during the war, leading to the name C. trigynus here being lectotypified with the K specimen. Eriophorum montevidense is here explicitly synonymised with C. trigynus as both species were very likely described from the same material at B.

CONCLUSION
After a long period of taxonomic uncertainty, by integrating recent molecular phylogenetic data with additional evidence from morphology and anatomy, Androtrichum is combined into Cyperus. As a result of this taxonomic change, a single monophyletic genus Cyperus is now recognised in the Cyperus Clade of tribe Cypereae.