(3146) Proposal to reject the name Xuris ( Iridaceae)

(3146) Xuris Adans., Fam. Pl. 2: (20). Jul–Aug 1763 [Angiosp.: Irid.], nom. utique rej. prop. Typus: non designatus. By the second half of the 18th century, several authors contemporary with Linnaeus published contrasting taxonomic arrangements of iridaceous groups. Philip Miller, Michel Adanson and Friedrich K. Medikus were perhaps the most prominent figures to disregard Linnaeus's (Sp. Pl.: 38–40. 1753) treatment of irises, which included 18 species in Iris L. and 1 species in Ixia L. (l.c.: 36). Conversely, they generated multigeneric classifications based mostly on the characteristics of the root system (bulbs or corms vs rhizomes), leaf morphology (equitant vs dorsiventral, basal rosette vs alternate or verticillate on stems), inflorescence (solitary terminal vs spicate), flower whorls (equal vs unequal), stamens (free vs fused) and stigmas (linear-cylindrical vs petaloid flattened) of the different groups of irises. Furthermore, those arrangements were intended to be respectful of names previously used by pre-Linnaean authors, such as Bauhin, Clusius, Dodoens or Tournefort, among others. Consequently, several generic names were validly published and available for use within the decades following publication of the Linnaean Species plantarum. However, many of these names were overlooked by subsequent authors describing new genera that overlapped entirely or partially with earlier ones. The present rejection proposal under Art. 56 of the Madrid Code (ICN; Turland & al. in Regnum Veg. 162. 2025) deals with one of those obsolete generic names, Xuris Adans. (Fam. Pl. 2: (20). 1763), which has remained almost forgotten for more than two centuries. However, as a validly published name under the ICN, it must be taken into account when competing with names of lesser priority. If Xuris is accepted at the genus rank, it has priority and therefore challenges some other well-established generic names such as Gattenhofia Medik. (in Hist. & Commentat. Acad. Elect. Sci. Theod.-Palat. 6: 418. 1790) (the “Vernae irises”) or Limniris (Tausch) Rchb. (Deut. Bot. Herb.-Buch: 43. 1841), nom. cons. (the “Laevigatae irises”), as shown below. Adanson (l.c.: (20)) applied the name Xuris to a group of unbearded irises with creeping rhizomes, six unequal tepals (the outer ones smooth, and the inner ones smaller but conspicuous), all fused into a short tube. These features distinguished it from the remaining genera he accepted in “Les Iris. Irides” (Iridaceae). But the most important point was the three synonyms that he connected to Xuris in the left column of a diagnostic row to be added to a general comparative table for the family (Adanson, l.c.: 60): (i) Dillenius, Hort. Eltham. 1: t. 155. 1732, referring to “Iris Am[ericana]. versicolor, stylo non crenato” (fig. 187) and “Iris Am. versicolor, stylo crenato (fig. 188)”; (ii) Plukenet, Phytogr.: t. 196, fig. 6. 1692, “Iris virginiana pumila, s[ive]. Chamaeiris verna, angustifolia, flore purpuro-caeruleo, odorato. D. Banister”; and (iii) Morison, Pl. Hist. Univ. 2: sect. 4, t, 6, fig. 11. 1680, corresponding to “‘Iris palustris lutea, s. Acorus adulterinus, I.B.”. Those cited figures were used by Linnaeus (l.c.) in the protologues of his Iris versicolor, I. verna and I. pseudacorus, respectively; albeit for the latter this is true only indirectly through a reference to his Hort. Cliff.: 19, No. 6. 1738. Notwithstanding, they undoubtedly represent Gattenhofia verna (L.) Medik. (in the case of Plukenet's), Limniris versicolor (L.) Rodion. (in the case of both Dillenius's) and L. pseudacorus (L.) Fuss (in the case of Morison's). Consequently, Xuris encompassed two different and very distant evolutionary lineages when described, and therefore it threatens both Gattenhofia (Iris subsect. Vernae Diels in Engler & Prantl, Nat. Pflanzenfam., ed. 2, 15a: 502. 1930) and Limniris (I. sect. Limniris Tausch, Hort. Canal.: [sine pag.]. 1823), nom. cons. Regarding Gattenhofia, it was described in detail by Medikus (l.c. 1790) to accommodate solely Iris verna L. into a multigeneric system (l.c. 1790: 414–443) together with other genera such as Iris L., Pseudo-iris Medik., Chamaeiris Medik., and Xiphion Mill. (Gard. Dict. Abr., ed. 4: [sine pag.]. 1754), to which he referred as the orthographic variant “Xyphion” (l.c. 1790: 418). Later, he completed his generic arrangement by adding both Biris Medik. (Philos. Bot. 2: 97. 1791a; in Staatswirthschaftl. Vorles. Churpfälz. Phys.-Ökon. Ges. Heidelberg 1: 257. 1791b), to accommodate only Iris sibirica L., and “Xeris Medik.” (l.c.: 1791a: 96; l.c. 1791b: 256), a name apparently not sufficiently characterised (nomen nudum) for an unidentified group of irises. Most of those generic names are currently available for use (see Crespo & al. in Phytotaxa 232: 1–78. 2015). The only exception to this is Pseudo-iris, a name rejected under Art. 14 against Limniris, which was conserved (see Applequist in Taxon 65: 1154. 2016) following the proposal by Crespo & Alonso (in Taxon 61: 684–685. 2012). Reichenbach (l.c.) based Limniris on Iris sect. Limniris Tausch (l.c.), a section first comprising the Linnaean I. pseudacorus and I. sibirica, which he broadened to include also two additional sections, I. sect. Xyridion Tausch (l.c.) and I. sect. Spathula Tausch (l.c.), both currently synonymised to Chamaeiris Medik. Rodionenko (in Bot. Zhurn. (Moscow & Leningrad) 92: 547–554. 2007) recovered the original sense of L. sect. Limniris, and lectotypified it on L. sibirica (L.) Fuss. Furthermore, Rodionenko (l.c.) expanded the genus to include other groups often referred to as “Apogon irises”, which he categorised into three subgenera and seven sections. Firstly, L. subg. Limniris with sect. Californicae (Diels) Rodion., sect. Laevigatae (Diels) Rodion., sect. Longipetalae (Diels) Rodion. (gen. Dielsiris M.B. Crespo & al., l.c. 2015: 64), and sect. Prismaticae (G.H.M. Lawr.) Rodion. Secondly, L. subg. Hexagonae with only sect. Hexagonae (Diels) Rodion. (gen. Phaeiris (Spach) M.B. Crespo & al., l.c. 2015: 58). And thirdly, L. subg. Ioniris with sect. Ioniris (Spach) Rodion. (gen. Joniris Klatt in Bot. Zeitung (Berlin) 30: 502. 1872) and sect. Chinenses (Diels) Rodion. (gen. Zhaoanthus M.B. Crespo & al., l.c. 2015: 59). A more detailed account of the nomenclatural history of the “Apogon irises” is shown in Crespo & al. (l.c. 2015: 28–33). Rodionenko's arrangement of Limniris s.l. is in full agreement with the “Limniris core” sensu Wilson (in Syst. Bot. 34: 277–284. 2009). However, it constitutes a morphologically variable aggregate in which up to five homogeneous groups can be recognised that, as noted above, may be regarded as distinct genera (Mavrodiev & al. in PLoS One 9(8)-e106459: 1–8. 2014). When circumscribed in a narrower sense, Limniris s.str. (L. sect. Limniris + sect. Laevigatae + sect. Prismaticae) includes ca 50 species plus 4 nothospecies, which are widespread in Eurasia, northern Africa and North America, albeit absent in the northernmost areas (Crespo & al., l.c. 2015). Although Limniris was nomenclaturally superfluous upon publication, as it encompassed Iris sibirica, the sole element included in Biris Medik. (l.c.: 1791a: 97; l.c. 1791b: 257), it is not illegitimate (Art. 52.4) as it has a basionym (I. sect. Limniris). The conservation of Limniris against Pseudo-iris is also effective against the earlier name Biris, since both Limniris and Biris are homotypic (Art. 14.4). But, although Limniris became correct after conservation against Pseudo-iris, it is still threatened by the earlier heterotypic Xuris (Art. 14.5). While threatening both Gattenhofia and Limniris, reinstatement of the obsolete and almost neglected Xuris would require the establishment of about 54 new combinations if displacing Limniris or 2 new combinations if displacing Gattenhofia. Both solutions would be highly disruptive, primarily for Limniris, as three of the four polynomials referenced by Adanson under Xuris belong here, so perhaps synonymisation with Limniris would be more logical. For those reasons, it appears most convenient to propose straightforward rejection of Xuris, rather than arbitrarily typify it on a species identifiable with either of the two aforementioned generic names. This approach would eliminate further unnecessary nomenclatural instability. In addition, at the time of submitting this proposal for publication (20 Jan 2026), the generic name Xuris is listed in the most widely consulted nomenclatural databases, such as IPNI, POWO, TROPICOS and WFO, and is also recorded in Index Nominum Genericorum (ING) (all sites accessed 15 Jan 2026). However, as far as we are aware, Xuris has never been used in general floras or monographic works dealing with irises, and species names are not available for the majority of taxa belonging to each of them. The only exception to this is maybe Rafinesque's (Fl. Tellur. 2: 14. 1837) reference to “Xuris”, a name which he reinterpreted and suggested to be a mistaken variant (sphalma) of the Linnaean Xyris (Sp. Pl.: 42. 1753): “Xuris mispelt [sic] Xyris by L. too near Iris”. In fact, Rafinesque (l.c.) established the new name Xuris fistulosa Raf. for an American plant currently known as Xyris baldwiniana Schult. (X. juncea Baldwin ex Elliot, nom. illeg.). This solution was later upheld by Rafinesque himself (Autik. Bot.: 189. 1840), who vindicated Xuris and Xyris to be merely orthographic variants. Indeed, Xyris is the type of Xyridaceae, the yellow-eyed grass family, in Poales (cf. APG in Bot. J. Linn. Soc. 181: 1–20. 2016). This commelinid genus encompasses about 280 species distributed widely across the tropical and subtropical regions of the world, mostly in South America, but extending northwards into Korea and the Atlantic territories of North America. It comprises annual or perennial, junciform herbs, exhibiting outstanding vegetative and floral features (see Kral in Kubitzki, Fam. Gen. Vasc. Pl. 4: 461–469. 1998) quite distinct from both the “Vernae” and the “Laevigatae irises” on which Adanson based Xuris. In fact, Adanson (l.c.: 546) referred Xyris of Linnaeus to another genus, Kotsjiletti Adans. (l.c.: 60, 544), that he adopted from Rheede (Fl. Malab. 9: t. 71. 1689). Despite Dandy's (in Regnum Veg. 51: 91. 1967) clarification of the application of both names, Xuris was sometimes reduced to synonymy with Xyris (POWO; accessed 20 Jan 2026), perhaps following the mistaken interpretation by Rafinesque (l.c.). In addition to the aforementioned points, it should be noted that the recovery of the obsolete and almost forgotten Xuris, which many botanists synonymise with Xyris, yields no practical purpose. In conclusion, for those who favour synonymisation of Gattenhofia and Limniris to Iris, this proposal has apparently no relevance, but it parallels the one recently submitted by the authors for rejection of Isis Tratt. (Crespo & al. in Taxon 74: 971–972. 2025). However, based on the evidence discussed above, we formally propose straightforward rejection of Xuris. This approach will best serve the stability of names by averting any potential disadvantageous and unnecessary nomenclatural changes. Should this proposal be rejected, from 2 to 54 new combinations would be required in Xuris, a generic name that, to the best of our knowledge, has remained almost forgotten and unused since its valid publication in 1763; this action would also create undesirable and unnecessary instability to nomenclature by relegating Gattenhofia or Limniris to synonymy. Conversely, acceptance of the present proposal could perhaps be the most effective way towards nomenclatural stability in the “Iris-flower clade”, and would avoid further confusion for those who, like the authors, favour a multigeneric treatment for the irises. We are very grateful, as always, to John Wiersema (Department of Botany, Smithsonian Institution, Washington, D.C., U.S.A.) and Melanie Schori (National Germplasm Research Laboratory, United States Department of Agriculture, Beltsville, MD, U.S.A.) for their editorial comments and useful suggestions.