Typification of eight current and seven related names and a new section in the genus Bromus (Bromeae, Pooideae, Poaceae)

Abstract During our nomenclatural revision of the genus Bromus L. (Poaceae) for Flora Iberica, we found that several names were still untypified: nine names in current use or their basionyms and five synonyms. Typifications are still needed since stabilising the names will facilitate their use. We propose lectotypes for Bromusalopecuros Poir., B.contortus Desf. (and the superfluous B.alopecuroides Poir.), B.benekenii (Lange) Trimen, B.intermediusGuss.subsp.divaricatus Bonnier & Layens, B.molliformis J.Lloyd ex Billot, B.lepidus Holmb., B.lepidus f. lasiolepis Holmb., Bromussubg.Stenobromus (Griseb.) Hack. and Bromussect.Stenobromus Griseb. Neotypes for B.erectus Huds. and B.ramosus Huds. and an epitype for B.intermediussubsp.divaricatus Bonnier & Layens are proposed. In addition, we identify an isoneotype for B.erectus and isolectotypes for B.lepidus and B.lepidusf.lasiolepis. The area inhabited by the typified taxa includes both Africa and Europe. All the selected types are in agreement with the current use of the names and, thus, our selections contribute to stabilising the nomenclature of the genus Bromus. A discussion is provided to justify the selections. In addition, we typified two supraspecific names B.subg.Stenobromus and B.sect.Stenobromus. Finally, a new section, B.sect.Penicillius Llamas & Acedo, is described.


Introduction
The genus Bromus L. (Poaceae) includes about 200 species distributed worldwide, with the greatest diversity and most complex taxonomy in south-eastern Europe and western Asia (Acedo and Llamas 2001). The taxonomy and nomenclature of this genus is difficult and the appropriate ranks of various supraspecific, specific and infraspecific taxa still remain uncertain and contested. In addition, sometimes its sections are raised to genera: Anisantha K.Koch, Boissiera Hochst. ex Steud., Bromopsis (Dumort.) Fourr., Bromus L., Ceratochloa P.Beauv., Nevskiella V.I.Krecz. & Vved. etc. Some authors raise the sections to subgenera (e.g. Stebbins 1981, Acedo andLlamas 1999).
In step with research carried out by various authors in the last decade on the taxonomy and nomenclature of the genus Bromus L., we here present a nomenclatural paper concerning the 8 names of well-known and floristically or coenologically important taxa belonging to Bromus that remain untypified at present and others related with them.
Besides the basic interest of the typification of the untypified names to stabilise nomenclature, it is valuable for any Flora to include type information and references to the exact places where the designation of types were published. All typifications in our work affect native and naturalised taxa occurring in the Iberian Peninsula, most of them also being present in other European countries and North Africa or have been established as aliens in many territories around the world. In any case, most of these names are applied widely due to their current distribution. We are applying Bromus sensu lato circumscription, since there is not sufficient data to split it into different genera.

Materials and methods
This study is based on analysis of relevant literature (every protologue and location indications included) and search for specimens or images of the following herbaria to identify original material: B, BM, BRI, C, G, FI, H, GOET, L, LD, LE, LEB, LINN, K, MPU, P, PH, PI, S, UPS and W (acronyms according to Thiers 2018+). Finally, by studying digital images or specimens, we designate the most suitable type in each case. All our decisions on typifications follow the rules and recommendations of the International Code of Nomenclature for algae, fungi and plants (ICN; Turland et al. 2018).
The references are consulted in the Biblioteca Digital del Real Jardín Botánico de Madrid (2018) at http://bibdigital. rjb. csic.es/ing/index.php, BHL (2017), Biodiversity Heritage Library at https://www.biodiversitylibrary.org/, Botanicus Digital Peter H. Raven Library Missouri Botanical Garden at http://www.botanicus.org/ and Gallica https://gallica.bnf.fr. All available images of specimens can be examined via JS-TOR Global Plants (2000 https://plants.jstor.org/ and many on the servers of several of cited herbaria. Currently accepted names are listed in alphabetical order, including their synonyms in each entry. Accepted names are in italic-bold, while junior synonyms are in italic-non-bold. Specimens seen are marked "!", images of specimens seen as "image!".  Poiret (1789: 100) stated in the protologue diagnosis "Panicula conferta erecta, spiculis oblongis subsessilibus, aristis inferne spiraliter contortis." followed by a description in French and the locotypic indication indicating this species grows around La Calle [now El Kala, El Tarf province, Algeria].

Bromus alopecuros
During our search for original material in the herbaria conserving the plants of Poiret (FI, H, P and UPS), we found only one specimen collected by him. There is a sheet in P (P02622864) from Numidia (Algeria "Numidia" included in the full title of his publication), registered as original material, bearing a single plant annotated as B. alopecuros with five labels transcribed below; one of them handwritten by Poiret:  Poiret (1789). In his diagnosis, Desfontaines adds that the spikelets are "quindecimfloribus, …pubescentibus". At the end, he has doubts about his plant and  transcribes a reference to the description of Poiret "An Bromus alopecuros? Poiret. Itin. 2. P. 100)". Subsequently, he includes a more detailed description and asserts "Habitat prope La Calle". The only known original material for B. contortus is the illustration quoted in Desfontaines (1798: plate 25) and a single sheet in P (MNHN-P P00320328) bearing two specimens mounted on it with a printed label: Herbier de la FLORE ATAN-TIQUE donné au Museum par M. DESFONTAINES and annotated "Bromus contortus" handwritten (unknown by whom). There is another label, handwritten by Desfontaines, containing the exact diagnosis and description as it appears in Flora Atlantica (Desfontaines 1798: 95). Maybe Desfontaines is surprised one specimen is very similar to B. alopecuros, but not the other one, justifying his doubt in the description where he states "perhaps B. alopecuros?" This sheet bears two plants. The one on the left seems to be Bromus lanceolatus Roth and the one on the right is a fragment (contracted and erect panicle with subsessile spikelets) of a specimen matching B. contortus Desf., that also exemplifies the typical resemblance to B. alopecuros Poir. Therefore, the sheet in question does not represent a specimen as defined in the Code (Turland et al. 2018); but each of the two plants on the sheet is a specimen in its own right. Only one specimen is original material for B. contortus Desf. It is possible that the illustration in Desfontaines (1798: plate 25) was drawn from the plant designated here as lectotype.
Choosing the specimen on the right of the sheet MNHN-P P00320328 ( Figure 2) as lectotype, which taxonomically matches B. alopecuros Poir., the name becomes a taxonomic or heterotypic synonym (Turland et al. 2018) of it, as Persoon (1805, 1: 95) asserted. Moreover, it is also possible that the heterogeneous material in this sheet is the origin of some misidentifications of B. lanceolatus Roth as B. contortus Desf.
Later, Poiret (1810: 703) describes Bromus alopecuroides "Bromus panicula conferta, erecta; spiculis oblongis, pubescentibus, quindecimfloris, subsessilibus; aristis infernè spiraliter contortis". This description is almost identical to the former of B. alopecuros. Its only difference is to include "pubescentibus, quindecimfloris", the same features Desfontaines (1798: 95) uses to describe B. contortus. Poiret continues adding the references to B. alopecuros Poir. and the synonym B. contortus Desf. Therefore, this name does not need a Lectoype as it is a superfluous and illegitimate renaming (Art. 52. 1) of Bromus contortus Poir. and has the same type we select here for that name.  in Flora Danica. In the original publication, Lange (1871: 5) describes a perennial Bromus living in forests, with nodding panicle.
An exhaustive search to find the material Lange (1871) cites as models for the illustrations in table 2826 (figs 1, 2), finally had a result. The two syntypes are conserved in C in the Flora Danica subherbarium, that contains specimens drawn in the magnificent work "Flora Danica" (Olof Ryding pers. comm. 2018). Their labels state "insula Lolland legit cl. E. Rostrup" (C10021728)" and "Jonstrup Vang, legit cl. H. Mortensen" (C10021729). Each folder indicates in handwriting that the specimens were drawn for "Flora Danica".
We choose as lectotype the latter sheet since its spikelets conserve most of its florets. The other sheet is in a more advanced phenological state and conserves, in most of its spikelets, only the glumes. The Rostrup specimen still conserves its basal part, with leaves, that is missing in the lectotype.  Scholz (1999: 436). Although both publications fulfil the conditions for a formal typification of a lectotype or a neotype, the Acedo and Llamas (1999: 73) publication appeared in February and is probably the first typification (ICN, Art. 10. 5). There are two important facts to consider. First, according to Kerguélen (1975: 100), there is original material in GOET. We consulted that herbarium and found two sheets that Schrader sent to GFW Meyer, which are indeed original material. Both bear handwritten labels by Schrader. The first sheet label says "B. commutatus Fl. Germ. Göttingen" and the second sheet label "B. commutatus Fl. Germ. var. spic. paulo brevior. Göttingen".  (Figure 4) Remarks. As a fire in his house destroyed Hudson's personal herbarium, most specimens were lost and only those borrowed by other botanists are extant. Some specimens which Hudson sent to Linnaeus are preserved in the herbarium of the Linnean Society of Lon-

don (LINN) but none of them is Bromus erectus
Huds. There is also one sheet conserved in BM that does not seem to be original material. As Hudson did not give any other reference and, as no original material of B. erectus Huds. is available, a neotype must be designated (ICN, Art. 9.13). We searched for material coming from Kent ("Cantium", England) in K and selected one specimen consistent with the protologue as neotype. It was collected in Kent, near Wye. There is a duplicate of it in BRI, which is an isoneotype. Remarks. Bromus hordeaceus L. is a very variable species with a complex nomenclatural history at subspecies level. The origin of many problems in the group is the name B. molliformis J.Lloyd, Fl. Loire-Inf. (Lloyd 1844: 314-315), which was invalidly published as a provisional name "je proposerais, si c'était une espèce nouvelle, de l'appeler Br. molliformis". The name was later validated as Bromus molliformis J.Lloyd ex Billot, Fl. Gall. & Germ. Exsicc. (Haguenau) 2: Cent. 14. 1854. Billot supplies a brief description for "cette espèce donnée par M.Lloyd dans sa Flore de la Loire-Inférieure sous le nom de B. divaricatitus Rohde?" Billot (Feb. 1854: 297-298) neglects to mention several details of Lloyd's description of B. molliformis and describes the species as follows: "Racine fibreuse. Chaume de 2-4 décimètres. Feuilles et graines inférieures mollement poilues. Panicule oblongue, droite, étalée, resserrée après la floraison; pédoncules courts, simples. Epillets oblongs, étalés, velus; arête égalant la glumelle, d'abord droite, à la fin tortillée divariquée, insérée à 1 1/2 millimètre du sommet obtus, échancré". Billot (1854) publishes the first validating description for Bromus molliformis. He includes the reference of the features to differentiate the new species from the closely related B. hordeaceus L. and from B. divaricatus Rhode. He sold his exsiccata to several herbaria. There is not a register of those herbaria. Currently, some of them are included in P where some original specimens with its species number (1386)   [sub. B. molliformis Lloyd]. They also mention Billot's exsiccate n. 1586; and include his new species in the complex of species having "arête … tordue sur elle-même et divariquée". Therefore, this name does not need a Lectoype as it is a superfluous and illegitimate renaming (Art. 52.1) of Bromus molliformis J.Lloyd ex Billot and has the same type we select here for that name. Bonnier and Layens (1894: 369) Kerguélen (1981: 27) combines it to Bromus hordeaceus subsp divaricatus (Bonnier & Layens) Kerguélen. Bonnier and Layens type material is unknown (Stafleu and Cowan 1976). Kerguélen (1975: 104) mentions Lloyd's type material is in herbarium NTM, pointing out two localities "Pornic, Saint Brevin" indicated by Lloyd. Nevertheless, no original material is preserved in NTM (Mary Laury Guerin, Com. pers. 2018).
As Bromus intermedius subsp. divaricatus Bonnier & Layens is an untypified name, we choose, as lectotype, the figure in Bonnier and Layens (1894: 369) that is definitely original material. However, as is common with figures, it is difficult to observe some diagnostic characteristics. Thus, we select, as epitype, a specimen collected by Lloyd, conserved in BM, consistent with the protologue. It is a sheet bearing two collections. We choose the specimen on the left from Pornichet (barcode BM001067302).
Bromus lepidus Holmb., Bot. Not. 1924: 326 (1924 Bromus lepidus Holmb., Bot. Not. 1924: 326 (1924 (Figure 8) Remarks. This taxon has a long history full of nomenclatural problems. Duval-Jouve (1865: 208), who only lists, without naming them, the variations in spikelet size and hairiness of Bromus species, e.g. Bromus mollis "microstachys glabre" and Bromus mollis "microstachys pubescent". This is the first mention of this taxon. Later Rouy (1913: 236) proposed a named variety under the genus Serrafalcus Parl.: S. mollis ß microstachys Rouy, giving "Bromus microstachys Duval-Jouve" (1865: 207) as a synonym and adding a diagnosis. Therefore, there is no doubt this is the same plant cited by Duval-Jouve (l. c.). Afterwards, Krösche (1924: 329) described B. gracilis Krösche, which unfortunately is a posterior homonym as Weigel (1772: 15) previously proposed this name for a different plant. Finally, Holmberg (1924: 326) solves this unfortunate situation and names simultaneously the new taxa Bromus lepidus Holmb. and B. lepidus f. lasiolepis Holmb. The herbaria having Holmberg material are K, LD and S. There are 13 sheets in LD collected before 1924, five in S and three in K. It is reasonably certain that Holmberg studied all those plants before the description of the species and form and that all of them are original material. Therefore, we decided to limit our choices to sheets with the annotations "B. lepidus mihi" and "B. lepidus f. lasiolepis Holmb.", as this annotation indicates that Holmberg is interpreting them as the new taxa he is going to describe. Remarks. Hudson (1762: 40) proposed the name Bromus ramosus for the wood Brome-grass from England. The protologue is a short diagnosis "BROMUS panicula ramosa nutante scabra, spiculis linearibus decemfloris arista longioribus, foliis scabris" followed by three polynomials as "synonyms", but without indication of the geographical area in which the new species lives, except an indication to "Anglia". As pointed out before, the fire destroyed Hudson's house as well as his personal herbarium. We were not able to find any sheet that could be "original material". Accordingly, we select a neotype of Bromus ramosus Huds. An annotation label indicates that Carmen Fraile verified and chose it as a neotype in 1994, but this designation was never published (ICN, Art. 7. 10). As this sheet agrees with the protologue, we accept her choice and make it effective here.