Typification and taxonomic remarks on five species names in Cytisus (Fabaceae)

Abstract This paper deals with the typification and taxonomy of five Mediterranean Cytisus species. Cytisus affinis, C. candidus, and C. spinescens nom. illeg., non Sieber ex Spreng. were described from Sicily by Karel Bořivoj Presl, Cytisus spinescens was described from Apulia (southern Italy) by Curt Polycarp Joachim Sprengel, and C. villosus was described from southern France by Pierre André Pourret (1788). Lectotypes are here designated for Presl and Sprengel’s names. A neotype is designated for C. villosus. The taxonomic revision of these five names confirmed that C. villosus Pourr. (= Cytisus affinis C.Presl) is the name to be used for the species occurring in the large part of the Mediterranean countries. Cytisus spinescens Sieber ex Spreng. (≡ C. candidus C.Presl = C. spinescens C.Presl, nom. illeg.) is the correct name for the amphi-adriatic species occurring in peninsular Italy, and along the NE coast of the Adriatic Sea. This species does not occur in Sicily and reference to this latter region in the protologues of both C. spinescens C.Presl and C. candidus C.Presl is a misinterpretation due, possibly, to exchange of labels.


Introduction
The Italian vascular flora includes 17 native Cytisus L. species and subspecies (Bartolucci et al. 2018) belonging to seven sections (Cristofolini and Troia 2006), and C. striatus (Hill) Rothm., a naturalised alien in Liguria . Half of these taxa are widespread in the Mediterranean region and occur in a large portion of Italy (e.g., C. hirsutus L. and C. villosus Pourr.); other taxa show a limited distribution and occur only in a few Italian regions (e.g., C. pseudoprocumbens Markgr.), or are narrow endemics (e.g., Cytisus aeolicus Guss. confined to the Aeolian Islands, Conte et al. 1998).
Several names in Cytisus, published during the 19 th and 20 th centuries, still lack a nomenclatural type and there are even doubts about the taxonomic position for some of these names (Peruzzi et al. 2015(Peruzzi et al. , 2019. Among them, there are three species described from Sicily by Karel Bořivoj Presl (1794-1852, standard botanical form C.Presl from Carl, Carel or Carolus) from Sicily, namely C. affinis C.Presl, C. candidus C.Presl, and C. spinescens C.Presl. These taxa were described only very briefly, in the form of footnotes within a list of taxa occurring in Sicily (Presl 1826: XIX). No locality was specified in the protologues. These names, as well as the related ones C. spinescens Sieber ex Spreng. and C. villosus Pourr., are typified here and their taxonomic relationship is discussed.
This contribution is part of the large project aimed at typifying all taxa described from Italy and recognising their loci classici in order to serve as a basis for further taxonomic studies (Domina et al. 2012;Passalacqua et al. 2014;Peruzzi et al. 2015Peruzzi et al. , 2019Brundu et al. 2017).

Material and methods
We performed a survey of the original material in the herbaria PR (National Museum, Prague) and PRC (Charles University, Prague) (acronyms according to Thiers 2019+), hosting the Presl's Sicilian collections (Stafleu and Cowan 1983). Further material has been searched in the main Italian and European herbaria that could host duplicates and/or the original material of C. spinescens Sieber ex Spreng. and C. villosus Pourr.: B, BM, BOLO, FI, G, K, MA, MAF-POURRET, NAP, P, PAD, PAL, RO, W, and WU. The articles of the International Code of Nomenclature for algae, fungi, and plants (herafter ICN) cited through the text follow Turland et al. (2018).  (Maiwald 1904: 180;Skočdopolová 1995). At the beginning of his career, he taught economic botany in the garden of Count of Malabaila de Canal (from 1826), later at the Faculty of Medicine (from 1829) and Philosophy (from 1833). As noted by Skočdopolová (1995), Presl frequently transferred herbarium specimens from Museum collections to his office at the university because of more suitable conditions for his work. For this reason, K.B. Presl's collections, including types, are variously distributed between today's herbaria PR and PRC. After a detailed search for original material of the above mentioned names, we found seven specimens deposited in PR and PRC putatively belonging to different gatherings and identified as three distinct taxa. We found two specimens of C. affinis C.Presl (PRC 450903 [ Fig. 1A], PR 375413 [ Fig. 1B]) collected by the author in Sicily in 1817. The specimen deposited in PRC (Fig. 1A) bears a Presl's label encompassing the species name and rather detailed locality, all written in italics, typical for his own collection (Domina and Štěpánek 2009). The specimen in PR (Fig. 1B) bears a label cut out from a specimen folder used at that time in C.M. Sternberg's herbarium, including the species name (at varietal rank, "Cytisus triflorus L'her. β. C. affinis Presl."), locality, collector and a short diagnosis against C. triflorus L'Hér. In addition, there is attached a small label from Presl's exsiccata collection "Flora sicula", suggesting that duplicates of this collection were distributed in the past and can be found elsewhere. Both specimens look very similar in respect of phenology and form of preparation and although they differ in the month of collection (May versus April), this likely originates from labelling of specimens in different times, and both specimens could be part of a single gathering. Both specimens are original material. They bear the name "Cytisus affinis Presl" written by himself, and in this case it seems unquestionable that the name C. affinis is based on specimens collected by Presl in Sicily. In any case, bearing two different dates, we prudentially consider them as two different gatherings. As the specimen in PRC [ Fig. 1A] is more complete, we designate it as the lectotype of the name. From the morphological study of this specimen it is obvious that it agrees with the short original description, so that it can be stated that C. affinis C.Presl is a heterotypic synonym of C. villosus Pourr. Interestingly, in the PR label this taxon is subordinate to Cytisus triflorus L'Hér., and Presl himself later recognised C. affinis C.Presl as a synonym of C. triflorus [written without name's authority], a species currently accepted under the name C. villosus Pourr. (see below), in his unpublished and undated second volume of Flora Sicula.

Typification of the names
More problematic are the specimens belonging to the original material of Cytisus candidus and C. spinescens. We have found one specimen belonging to Cytisus candidus in PR ( In addition, in W there is a sheet (W333912 photo!) [ Fig. 2B] bearing four specimens with four labels bearing different names and collected in several localities of peninsular Italy: Cytisus spinosus DC. (two specimens from the Gargano), C. ramosissimus Ten. from the mountains near Castellammare, and C. nanus Willd. from Calabria. Although all specimens from PR and PRC bear Presl's handwritten identifications, the plants belong to the same taxon and all agree with the protologues of both Presl's C. candidus and C. spinescens. More specifically, both names were allegedly based on material originated from Sicily, as can be deduced from (i) descriptions of both taxa included in Flora Sicula (Presl 1826), and (ii) specification about the locality of C. candidus being collected in two carbonate promontories near Palermo ("Habitat in regione collina in saxosis apricis sterilibus ad promontorium Zafferana una vire, altera vire in monte Pellegrino", see Presl,undated,unpub. msc. Flora Sicula vol. 2) or in Sicily in general (Presl's annotations on two specimens deposited in PRC "E[x] Sicilia"; PRC 454917 and PRC 450971). Importantly, from the taxonomic point of view, both Cytisus candidus and C. spinescens C.Presl, are heterotypic synonyms of C. spinescens Sieber ex Spreng. (see below), a taxon which, besides Presl's records from Flora Sicula, has never been reported from Sicily (Bartolucci et al. 2018). In fact, C. spinescens Sieber ex Spreng. is a taxon confined solely to the Italian peninsula (northwards to Latium, Umbria and Marche) and to the NE coast of the Adriatic Sea. In addition to the doubtful location (Sicily), it has become obvious from the elements specified below that these specimens were not collected by Presl himself, but by Franz Wilhelm Sieber (1789-1844) and by Charles-François-Louis-Alexandre [Luigi] Thomas (1784-1823) (cf. Burdet 1978; see also an annotation to the Table 1), respectively. We hypothesise that these discrepancies in locations and collectors have likely originated from dividing and postponing the labelling of these specimens by Presl himself. Such a mistake has previously been documented in Asplenium lepidum C.Presl, which was allegedly collected by him in Bohemia, but actually by Anton Rochel (1770-1847) in the region of Banat (currently in Romania and Serbia) (P. Mráz, unpublished data).
In the case of Cytisus candidus, the specimen PR 375660 (Fig. 1C) bears, in addition to Presl's label, also Sieber's original label of "Cytisus spinosus DC." from his exsiccata collection "Plantae Neapolitanae et Apulae". As stated on both labels, it was collected in Gargano, where this species (currently C. spinescens Sieber ex Spreng.) occurs (Fenaroli 1970;Bartolucci et al. 2018). Interestingly, Sieber's original label is missing in the specimen found in PRC (PRC 454917, Fig. 1D), which bears two labels written by Presl only (Table 1). We found Sieber's duplicates of this gathering also in JE (JE 00021324 Photo!) and W (W 333912 photo!, plant on the left bottom, Fig.  2B). Importantly, both these specimens bear Sieber's exsiccata labels and the plants show the same colour and character as the specimens housed at PR (Fig. 1C) and PRC (Fig. 1D). We here selected the specimen at PR (PR 375660), bearing the original label from Sieber's "Plantae Neapolitanae et Apulae" collection, as the lectotype of C. candidus C.Presl. The specimen PRC 454917, as well as the duplicates in JE and W, are therefore isolectotypes of C. candidus C.Presl.
The three remaining specimens (PR 375417 [ Fig. 2A], PRC 450971 [ Fig. 2C], and PRC 452282 [ Fig. 2D]) are again morphologically very homogeneous and were consistently identified by Presl as "Cytisus spinescens Presl", although labelled as being collected from three different sites (see Table 1). Very important in this respect is the sheet W 333912 (Fig. 2B), with the specimen in the right bottom corner "Cytisus nanus Willd." collected by Thomas in Calabria. Importantly, a similar label showing the same plant name and locality accompanies the specimen PR 375417 ( Fig. 2A), whose plant shows similar / identical habitus as the one at W. The same can be argued for the specimens from PRC (PRC 450971 [Fig. 2C], and PRC 452282 [ Fig. 2D]), albeit missing Thomas' label. On the contrary, one of the PRC specimens (PRC 452282)  2D) and was presumably written in 1832 or later. Reference to Sieber's collection from Capri, also noted on specimen PR 375417 ( Fig. 2A). Reference to C. nanus was probably wrongly ascribed to Sieber and, in fact, it refers to the specimen of L. Thomas "Cytisus candidus Presl. / E Sicilia." Presl's handwritten label PR 375417 ( Fig.  2A) (Moret, 1993(Moret, , 1999, where was also active J.C. Schleicher. Gathering collected by one of Thomas was most probably sent to Prague by Schleicher (reference to Sprengel's publication given) "Cytisus nanus Willd / seu nova species / Calabre" Handwritten label probably by Ch.F.L.A. Thomas, but not entirely sure if written by him or by his brother A.L.E. Thomas. Based on the note on duplicate specimen kept in Wien (W 333912, Fig. 2B). According to Burdet (1978), the label is more probably written by A.L.E. Thomas, although presumably collected by Ch.F.L.A. Thomas, who worked in Calabria  1D), we hypothesise that Sieber's label attached to the specimen of C. spinescens C.Presl (PRC 452282, Fig. 2D) emerged from a mistake and was, in fact, exchanged with that of C. candidus (PRC 454917, Fig. 1D). Since the specimen PR 375417 [ Fig. 2A] contains the best preserved plant and bears both Presl's identification label and original label by Thomas, we designate it as the lectotype of the illegitimate name C. spinescens C.Presl. Consequently, we consider the specimens in PRC (Figs. 2C, 2D) and W (Fig.2B, the plant in the right bottom corner) as duplicates of Thomas' collection from Calabria, and hence isolectotypes. A possible scenario leading to the current "messy" state is as follows. During the work on his Flora Sicula (between 1817 and 1825-1826), K. B. Presl had access to five gatherings of Cytisus from Italy. Two of them (C. affinis C.Presl) were part of his own herbarium and were collected by him in Sicily, another two (one by Sieber, one by Thomas) were part of Prague National Museum collections and came from Italian mainland. The fifth is a Sieber's gathering from Capri Island (Campania, southern Italy) and bears Presl's annotation 'Sieber'. It is deposited in PRC (PRC 455779) with no duplicate in PR. Our hypothesis is that Presl divided the museum specimens and transferred fragments to his own herbarium, and vice versa, donating duplicates of his own collection from Sicily to the Museum. We suppose also that during this "fragmentation" of specimens, he probably did not annotate carefully these fragments, and this may be the main reason for the chaotic situation concerning these collections.
Based on the morphology of the specimens of Presl's C. candidus and C. spinescens, which agrees with the short original descriptions, we conclude that both names are synonyms of C. spinescens Sieber ex Spreng. Because C. spinescens C.Presl was described about seven months later than C. spinescens Sieber ex Spreng. Cowan 1983, 1985), and because both names are based on different types (see also below), Presl's name is a later and heterotypic homonym of C. spinescens Sieber ex Spreng., illegitimate according to Art. 53.1 of the ICN. Consequently, Presl's name should not be used as an accepted name as it is currently treated in The Plant List (2019) or in the International Legume Database (Roskov et al. 2006) andin Euro+Med Plantbase (Euro+Med 2006). Concerning the name C. spinescens C.Presl, it is noteworthy that in his unpublished second volume of Flora Sicula, Presl wrote that its provenance was unclear for him ('locus specialis mihi amplius non constat') and unclear was for him also the status of C. spinescens Sieber ex Spreng. with respect to C. candidus ('Quid vero C. spinescens Spreng. … est … An species sequens?' [the next species in the manuscript is C. candidus]).  Fig. 2B]).

Typification of the names
Note. As Sprengel based his description on the exsiccata series collected and issued by F.W. Sieber, the best solution for typification would be to choose the specimen from Sieber's collection seen by Sprengel himself. Unfortunately, after the death of his son, Sprengel's rich herbarium was divided into many parts and sold in small portions to different specialists and institutions (Stafleu and Cowan 1985). The largest part, containing the collections of many botanists and among them also those by Sieber, was bought by B in 1890 (Urban 1891), and subsequently destroyed during World War II. We found unequivocal duplicates of this F.W.Sieber's collection in herbaria PR, W and JE, and as shown above also in PRC, although incorrectly labelled later by Presl. It is interesting to note that Presl based his later homonym C. spinescens on a different gathering (Thomas' collection), while he described C. candidus on a F.W.Sieber's gath-ering. As the above designated lectotype of C. candidus belongs, without any doubt, also to the original material of C. spinescens Sieber ex Spreng, we designate it also as the lectotype of the latter name. Cytisus candidus C.Presl thus becomes a homotypic synonym of the prioritary name C. spinescens Sieber ex Spreng.
This brings also another nomenclatural consequence: when treating C. spinescens Sieber ex Spreng. as a member of the separate genus Chamaecytisus Link, the correct name is Chamaecytisus spinescens Rothm. This is because Rothmaler (1944) based his intended "new combination" on Presl's illegitimate name, and thus accidentally published a replacement name (Art 58.1 of the ICN), which prevents making the combination based on legitimate Sprengel's epithethon. Note. We did not find any original material for this name either in MAF (MAF-Pourret collection), and P (the general collection and the special Pourret's collection named "Chloris narbonensis"), where Pierre André Pourret's (1754Pourret's ( -1818 collections are mainly kept (Stafleu and Cowan 1983), or in other relevant herbaria (BM, FI, MPU, and UPS; see Stafleu and Cowan 1983: 368). It seems, therefore, that the original material for this name is lost. This possibility is not surprising giving Pourret's dramatic escape from France to Spain in 1789 and his forced exile (Galibert 1856), followed by several war events (Stafleu and Cowan 1983: 368). Because the original material of C. villosus is lost, we have decided to choose a neotype represented by the specimen G00007761 housed at G-DC. This specimen has been previously selected by Cristofolini and Fumeaux (Cristofolini and Troia 2006) as the lectotype of C. triflorus L'Hér. [1791]; an illegitimate name (a later homonym of C. triflorus Lam.

Cytisus villosus
[1786]) being conspecific with C. villosus Pourr. (see Polhill 1978;Cristofolini and Troia 2006). Importantly, as Cytisus triflorus L'Hér. has been accepted as the conserved type for the generic name Cytisus Desf., nom. cons. (Appendix III of the ICN), it becomes automatically a homotypic synonym of Cytisus villosus Pourr. -which is the accepted name of the generitype of this genus.
to Alena Richterová (National Library, Praha) for her help in searching for Presl's unpublished manuscript. This work was supported by the "Progetto di Ricerca di Rilevante Interesse Nazionale" (PRIN) "PLAN.T.S. 2.0 -towards a renaissance of PLANt Taxonomy and Systematics" lead by the University of Pisa, under the grant number 2017JW4HZK (Principal Investigator: Lorenzo Peruzzi). Work of OŠ was supported by Ministry of Culture of the Czech Republic (DKRVO 2019-2023, that of PM by statutory funds of Charles University