Proposal to conserve the name Lycopsis pulla (Boraginaceae) with a conserved type

The name Nonea pulla has been in general, uncontroversial use for over two centuries to designate a perennial species of scant ornamental value and no economic importance, but noticeable by its unusual, almost black purple flowers and, being locally a rare and threatened plant, with an extensive floristic and conservational literature referring to it. Its area extends from Central Europe eastward throughout European and Asian Russia as well as to the Balkan countries, Anatolia, Caucasia, and the Iranian highlands; in the non-European part of its range it is represented by at least half a dozen closely related taxa now mostly treated as subspecies (see Riedl in Rechinger, Fl. Iran. 48: 247–248. 1967; Čerepanov, Svod Dopol. Izmen. Fl. SSSR: 115. 1973; Baytop in Davis, Fl. Turkey 6: 410. 1978) or, at least in part, assigned to separate species (Selvi & Bigazzi in Pl. Syst. Evol. 227: 1–26. 2001). Continued use of the name Nonea pulla for the species in question is in jeopardy. Trouble started with a paper by López & Jarvis (in Anales Jard. Bot. Madrid 40: 341–344. 1984) pointing to the fact that the name Lycopsis pulla, the accepted basionym of Nonea pulla, was not first published in 1759 by Linnaeus (Syst. Nat., ed. 10: 916) as had been generally assumed, but in 1758 in an index entry to Loefling (l.c.), validated by Loefling’s phrase name Lycopsis procumbens, foliis integerrimis, calycibus fructuum inflatis and associated descriptive matter in Swedish and Latin, on pp. 66–67 and 81. López & Jarvis (l.c.: 343) correctly stated that the type of the 1758 name must be selected from among Loefling’s Spanish material on which the validating descriptions were based; this material does not belong to the species generally known as Nonea pulla, which is absent from Spain. They further implied, to our mind correctly, that the author of the index with the 1758 binomial must be Linnaeus (the editor of the book) as it cannot logically be Loefling (the author of the posthumously published manuscript); and they assumed that subsequent usage of the name L. pulla by Linnaeus in his own works is just that, a later (mis)usage of the 1758 version, not a different name with a standing of its own. By consequence, they suggested that the combination Nonea pulla (L.) DC. (in Lamarck & Candolle, Fl. Franç., ed. 3, 3: 626. 1805) be used for the Spanish species known at that time as N. ventricosa (Sm.) Griseb. (and currently as N. echioides (L.) Roem. & Schult.)—a switch of usage that would have caused fatal confusion had it been followed. Later that year, Greuter & al. (Med-Checklist 1: 103. 1984) concluded differently on two points. First, Lycopsis pulla 1758 (irrespective of authorship) and Lycopsis pulla 1759 are two different names, each with its own type, the latter being an illegitimate later homonym. Second, Candolle’s Nonea pulla is, by implication, based on the later homonym and is therefore a legitimate replacement name with priority from 1805. The first point is supported by a number of facts: (1) Linnaeus, under L. pulla, makes no reference to Loefling’s publication either in 1759 or in any later publication; (2) the respective diagnostic phrases, while generally similar, differ in one point of note: instead of procumbens, as in 1758, in 1759 it has caule erecto; (3) in 1762, Linnaeus (Sp. Pl., ed. 2: 198), in the first expanded treatment of L. pulla, confirms the diagnostic value he attributes to habit (“Differt a L. vesicaria caule erecto”); and (4), again in 1762, the known distribution of L. pulla is given as “Tataria, Germania”, without mention of Spain. The second point is also borne out by the preceding consideration (1): since L. pulla 1758 was utterly ignored by botanical authors, reference to “lycopsis pulla, Linn.”, as by Candolle in 1805 under his new N. pulla, can only mean the later homonym of 1759. Kazmi (in J. Arnold Arbor. 52: 676. 1971) designated Herb. Linn. 190.2 (LINN) as the type of L. pulla 1759, a designation accepted by Jarvis (Order Out Of Chaos: 648. 2007); the lectotype (López & Jarvis, l.c.) of L. pulla 1758 is Herb. Linn. 190.6 (LINN), collected by Loefling in Spain; it is a likely duplicate of the lectotype of L. echioides L. (Sp. Pl., ed. 2: 781. 1762), Herb. Linn. No. 71.7 (S), previously designated by Edmondson (in Willdenowia 8: 23. 1977). Jarvis (l.c.), in a comment under Lycopsis pulla 1759, mentioned the earlier homonym of 1758 (not accounted for by the Linnaean Plant Name Typification Project because it appears in a work not authored by Linnaeus) and added: “a conservation proposal may be necessary to assure its use [of N. pulla] in the traditional sense”. Under the foregoing scenario, such a proposal would not be required, Nonea pulla DC. being available for use and preventing the transfer of L. pulla 1758 to Nonea. But there is a “but”. In 1800 Bernhardi (Syst. Verz.: 127), when transferring “Lycopsis pulla L.” [1759] to Nonea (‘Nonnea’), renamed it N. erecta Bernh.—a straightforward replacement name, without descriptive matter. That name would normally be illegitimate, but owing to the homonymous status of the replaced synonym it is legitimate and, as it predates N. pulla DC., it causes illegitimacy of the latter. Kerguélen (Index Syn. Fl. France: 123. 1993) was the first to take up the name N. erecta, but few took note initially. Acceptance of N. erecta as the correct name for our species, probably prompted by its adoption in Tela Botanica’s eFlore (http://www.tela-botanica.org/page:accueil_botanique), is currently incipient in a few countries of Central Europe. Among the examples known to us the following are of note: Aeschimann & Heitz (Index Syn. Fl. Suisse: 149. 1996), Lauber & Wagner (Fl. Helvet.: 120. 1996), Rothmaler (Exkurs.-Fl. Deutschl. 1, ed. 19: 388; 4, ed. 10: 557. 2005), and Tison & al. (Fl. France Médit.: 1295. 2014). This weighs little as yet, when compared to the compact and uniform usage of Version of Record (identical to print version). 1133 TAXON 63 (5) • October 2014: 1132–1133 Cecchi & al. • (2321) Conserve Lycopsis pulla N. pulla throughout the 19th and 20th century. It is not too late, in our opinion, to conserve the latter name. The question is how best to proceed. Conserving Nonea pulla DC. itself in the sense of its intended but illegitimate basionym and against the homotypic N. erecta Bernh. would be a definite option, but has the disadvantage of late-starting priority of the conserved name. Our initial idea was to have Lycopsis pulla 1759 conserved against L. pulla 1758, but apparently this would not work. Due to what we believe is an unintentional oversight, Art. 6.4 of the Melbourne Code (McNeill & al. in Regnum Veg. 154. 2012) does not restore legitimacy of a name when its basionym is conserved (only when the name itself is conserved); therefore, even if legitimacy of L. pulla were restored by conservation, N. pulla, perhaps nonsensically, would remain illegitimate. The solution we offer above is a third way out— the most elegant we believe. By conserving L. pulla 1758 with the designated type of L. pulla 1759, as Art. 14.9 of the Melbourne Code (l.c.) permits, the latter becomes eo ipso an isonym of the former. For the purposes of nomenclature such isonyms have no existence, and any reference to a later isonym (at least prior to 1953) is a correctable error. The effect is that N. erecta is an illegitimate name and N. pulla a legitimate combination based on L. pulla: all the complications previously described have vanished as by enchantment.