Nomenclatural revision of Delphiniumsubg.Consolida (DC.) Huth (Ranunculaceae)

Abstract Recent molecular phylogenetic studies have indicated that Aconitella is embedded in Consolida, which in turn is embedded in Delphinium. We choose not to split the genus Delphinium (c. 300 species), as it is horticulturally and pharmaceutically important, by conserving a broad Delphinium by transferring the names from Consolida and Aconitella to Delphinium s.lat., and more precisely in the resurrected D.subg.Consolida. Including 58 species of Aconitella and Consolida within Delphinium causes fewer nomenclatural overall changes than do alternative schemes because most of the species of Aconitella and Consolida were once named under the name Delphinium. We present here the list of synonyms for the species once named under Consolida or Aconitella and gather the information relative to the types of these names. Two new combinations are provided, and 21 lectotypes are designated here.

Based on a molecular phylogenetic study with a broad taxonomic sampling, Jabbour and Renner (2011) first found that Consolida and Aconitella were embedded in Delphinium. More precisely, Aconitella was nested within Consolida, which in turn was nested within Delphinium. More recent analyses confirmed this result (Jabbour and Renner 2012a;Wang et al. 2013;Xiang et al. 2017) but do not support the different subgroups (the "Untergruppen or Tribus") previously described in Consolida (Huth 1895). Thus, the overemphasis on distinctive characters (see Pfeil and Crisp 2005) in Consolida and Aconitella led to recognizing a paraphyletic Delphinium. As a consequence of these results, the last author previously decided (Christenhusz et al. 2018: 73) not to split the large genus Delphinium (c. 300 species), as it is horticulturally and pharmaceutically important (Tamura 1993), by conserving a broad Delphinium (see Kadereit et al. 2016)  As flower morphological characters support a clade including Consolida and Aconitella (Jabbour and Renner 2012b), we treat Consolida as a subgenus of Delphinium in this article. We re-introduce here Delphinium subg. Consolida (DC.) Huth and list its 58 species with information relative to their typification.

Methods
We analysed the original material cited in the protologue of each taxon and compiled the relevant synonyms. Except for rare cases, the infraspecific taxa of Consolida are not mentioned as taking taxonomic decisions at this taxonomic level is beyond the scope of this work.
Based on the methodology of typification followed by Al-Shehbaz and Barriera (2019), we provide notes about the typification, especially if a lectotype is designated here for the first time or if earlier lectotypifications were incomplete or erroneous. In some cases and some relatively recently described species, the holotype was not found in the mentioned herbarium. When we consider that further investigations are necessary, we have decided not to designate a lectotype. Accepted names are in bold italics and listed alphabetically. The reference of cited specimens (herbarium code, and when available, the barcode) are provided. The herbarium codes follow Thiers (2018). Specimens marked '!' were examined in the scope of this paper. Note that JE, WU, and W barcodes have temporary barcodes and are susceptible to change in the future (Jochen Müller, Dieter Reich, and Christian Braüchler, curators, pers. comm.).

Results
Delphinium subg. Consolida consists of 58 species. Two new combinations are made, and 21 lectotypes (including three second-step lectotypifications) are designated herein. Notes. Consolida ambigua auct. (non D. ambiguum L.) is a misapplied name of D. ajacis in most floras, such as "Flora Europaea" (Chater 1993) and "Flora of Turkey and the East Aegean Islands" (Davis 1965). For details, see Janchen (1965: 34). Notes. The misinterpretation of D. anthoroideum by Boissier in "Flora Orientalis" (1867) is clarified by Chowdhuri et al. (1958). Among the isolectotypes of D. anthoroideum, only G00390151 and P00195789 bear the date "1837". Notes. Chowdhuri et al. (1958) indicated the holotype at W, but we were unable to find it. No duplicate from the above herbaria was annotated by Huth, and we choose to designate WU 109667 as the lectotype for now. Notes. Iranshahr (1992) indicated that the holotype is kept at P, whereas Munz (1967a) indicated that it is kept at LE, but without having seen it. We found a duplicate at P and two at LE. None of them seem to have been annotated by Bunge. The P00197235 sheet bears three specimens and two handwritten labels. These two labels are in Latin, probably from Lehmann's hand, and correspond to the locality indicated in the protologue. They are almost identical (the left one carries "Delphinium sp ?" and "10 Sept.", and the right one bears the collection year 1841 and no identification). Both labels were stuck on Bunge's printed handwritten labels "Reliquiae Lehmannianae. Herb. Al. de Bunge.", and the right one is itself stuck to another label ("Rel. Lehm. N°.38.") written by a different hand, probably a curator of P. At LE, the labels bear the exact mention of the locality as in the protologue (in German) on a preprinted label "Alexandri Lahmann/ Reliquiae botanicae. Al. Bunge." and one of both bears the full date. Notes. When describing Delphinium camptocarpum, Ledebour quoted Karelin's gathering "ad latus orientale maris caspii" and described two varieties (D. camptocarpum var. dasycarpum and D. camptocarpum var. leiocarpum) according to the presence or not of indumentum on the follicle, but without citing any material. Several of Karelin's gatherings of D. camptocarpum were found at LE. Four of them belong to Ledebour's herbarium (two can be attributed to D. camptocarpum var. leiocarpum, but it is unclear for the other two) and can be regarded as type collection. We synonymise D. camptocarpum var. leiocarpum under the autonym.

Delphinium brevicorne
Nevski's indication (1937) that the type is housed at LE can be considered a firststep lectotypification, and we designate here the specimen LE00050875 as the secondstep lectotype. Notes. The specimen E00438700 is explicitly designated as the holotype in the protologue, whereas the indication of collection locality and date is only found on B 10 0264874 and GH00038199. Notes. In the protologue of D. flavum, Candolle (1817) described two varieties (var. velutinum and var. glabrum) based on the pubescence and the bracteole position on the pedicel. We synonymize here D. flavum var. velutinum with the autonym. Candolle indicates that he saw the specimen in Olivier's herbarium (now at P), whereas we found at G-DC a duplicate received in 1822. At P, we found two sheets of the Olivier and Bruguière's gathering. It is not sure whether Candolle annotated these sheets. The specimen P00197330 bears the label "dans les lieux incultes steriles de Bagdad a Kermancha" and a mixture of both varieties (the specimen in the bottom right corner corresponds to D. flavum var. glabrum). We designate as lectotype the six other specimens on this sheet. A single and fragmentary specimen of D. deserti, labelled "D. deserti" by Boissier, was found in G-BOIS (G00788308) and probably came from a P duplicate. We designate as lectotype of D. deserti the specimen P00197319, the only one bearing a priori an annotation from Boissier.

Delphinium glandulosum
Notes. The description was based on the specimen in G-BOIS, a folder containing three sheets. Only this specimen was annotated by Boissier, which is, therefore, the holotype. Notes. Delphinium orientale Gay was misapplied instead of D. hispanicum in most Floras e.g. "Flora Orientalis" (Boissier 1867), "Flore de l'Afrique du Nord" (Maire 1952), "Flora of Syria, Palestine and Sinaï" (Post 1932), "Flora of Turkey and the East Aegean Islands" (Davis 1965). See Greuter and Raus (1989) for a discussion of the nomenclature. Notes. Although Nevski annotated both duplicates at LE and K, the latter is clearly labelled "Dupla", and we considered LE00050815 as the holotype.

Notes. Boissier's annotation is found on P00198677 and G00788305, indicating
Boissier based the species description on these duplicates. Therefore, the lectotypification is justified. Notes. When describing new species, Bunge usually annotated the specimens cited with "mihi" or "m." behind the species name. Unfortunately, we did not find any annotation from Bunge on the different duplicates. P00198841 bears ten individuals and three different preprinted Lehmann's collection labels, all corresponding to collection number 36. One of them is from Lehmann's hand with the locality (in Latin) and the date indicated in the protologue. No date is indicated on the other duplicates. LE (photo seen at BM) bears a label with the locality translation in German. After an investigation by the curator, the LE specimen was not found. Notes. Candolle did not mention any specimen in his protologue when he described D. pubescens. Although some original material of the "Flore Française" can be at MPU or P (Stafleu 1967), where Candolle worked before moving to Geneva, we found a Candolle's gathering dated 1807 at G in the prodromus herbarium, that we designate as the lectotype. Notes. The holotype of D. pusillum bears a handwritten protologue by Labillardière. Munz (1967a) indicated that he saw an isotype of D. pusillum at K (obviously K000692380), but it is not clear whether this specimen belongs to the type collection or not. As explained by Chowdhuri et al. (1958), when dealing with D. oliganthum, the Haussknecht's gathering housed at G and K is a mixed collection of D. hellesponticum and D. oliganthum. In the Boissier herbarium, the folder of D. oliganthum contains two sheets, one with D. hellesponticum (G00788351), which bears the original label, and the other one with D. oliganthum (G00788351a), both annotated in 1956 by M. Hossain, one of the co-authors with P. K. Chowdhuri and P. H. Davis. At G, K and P duplicates also contain both species. We complete the first step of lectotypification by Chowdhuri et al. (1958), who indicated G and K simultaneously in designating the specimen on the sheet G00788351a at G-BOIS as the second-step lectotype.  Davis (1965) indicated that the holotype was kept at LD. Notes. In the protologue, Boissier quoted an unnumbered Haussknecht's collection with the following indication "in graminosis montium Syriae borealis prope Marasch alt. 4000'". Davis (1965) indicated as "types" a specimen housed at K. However, Boissier did not examine this specimen and based his description solely on specimens in his herbarium, where we found a collection folder containing two sheets. One bears a printed label of Haussknecht's collection with his handwritten additions of locality "In apris v. Karabigukle et pr. Marasch", and the date "Aug. 65". The other one bears three of Boissier's handwritten labels, one of which bears the mention of the locality "Montes azia/ Marasch" and the date "1865". At JE, K, and P, we found Haussknecht's duplicates (JE00018589, K000692378, and P00201052) with a label handwritten by himself bearing the locality "Uffoschikle" (or "Uffoschakli") with the date 11 July 1865. A duplicate at JE bears, in addition, the number 970. All these specimens correspond to syntypes. Notes. In the protologue, Boissier (1867) based the species description on the specimen Kotschy 884 housed at W ("in herb. Mus. Vindob!"). Iranshahr (1992) indicated to have seen this specimen at G. However, after investigation, the specimen Kotschy 884 was found neither at W, nor at WU, nor G. In G-BOIS, we found a Kotschy's gathering of D. teheranicum with a fragmentary specimen (probably a part of the specimen cited in the protologue) and a label written by Boissier. We designate this specimen as the lectotype. Notes. In his protologue, Boissier (1867) cited two gatherings: "in Bithynia, Thirke" and "circa Amasia et Tokat Anatoliae, Wiedem.". Wiedemann's gathering at G-BOIS is here chosen as lectotype. Munz (1967a) indicated that he saw a duplicate of this gathering at GH and K. However, the duplicate at GH could not be found by the curator. It is unclear whether the Wiedemann's gathering at K of D. thirkeanum (K000075588), which is apparently a duplicate from a specimen at LE, belongs to the type material. Moreover, Huth (1895) quotes a specimen at LE, which was not observed.  (Fig. 2D)).

Notes.
In his protologue, Boissier cited two gatherings of Aucher-Eloy: "N.76. Alep" and "77. Persia". Chowdhuri et al. (1958. 22: 418)  Notes. Halácsy based his description of D. tuntasianum on the specimen Tuntas 1245 from the "Plantae exsiccatae Florae Hellenicae" collection. Three Tuntas specimens of D. tuntasianum, each one including several plant individuals, were found at WU: Tuntas 1245, Tuntas 800, and Tuntas s.n. Examination of Tuntas 1245 shows two different dates: "23-28 Jun. 1910" (printed) and "10/23 Mai 1911" (handwritten), this later date being the same as the gathering "s.n.". Therefore, the gathering "1245" is probably a combination of two different collections. We designate it as a lectotype despite the uncertainty regarding its collection date. Notes. In his protologue, Huth quotes two syntypes (Pančic 1881 seen in three different herbaria, and Baldacci 1889 seen in the Barbey herbarium, now G). At G, there is a folder with two sheets of the second gathering that we designate here as lectotype.

Discussion and conclusions
According to the nomenclatural revision presented here, Delphinium subg. Consolida consists of 58 species. This work will facilitate a taxonomic study aimed at revising the circumscription of sections within D. subg. Consolida. Infrasubgeneric relationships were tackled and discussed in Jabbour and Renner (2011). Moreover, a thorough taxonomic study of the species-rich subgenera Delphinastrum (DC.) Peterm. and Oligophyllon Dimitrova is now timely, as they are the last subgenera in Delphinium that still require revision.
We thank three reviewers (Andrey Erst, Andriy Novikov, Martyn Rix) and the Editor (Marco Pellegrini) for their helpful comments on the manuscript. The MNHN gives access to the collections in the framework of the RECOLNAT national research infrastructure.