Panarctic Flora

672132 Draba glabella Pursh

• Pursh, Fl. Amer. Sept. 2: 434 (1813). Described from North America.
• ?Draba hirta L., Syst. Nat., ed. 10, 2: 1127 (1759). Type not designated (Jarvis 2007: 486). See notes.
• ?Draba norvegica Gunnerus, Fl. Norveg. 2: 106 (1776). Holotype: Norway. Oeder, Fl. Dan. 1, 3: t. 143 as Draba pyrenaica (?) Oeder. 1764. See notes below and to Draba rupestris.
• Draba daurica DC., Syst. Nat. 2: 350 (1821). Type (G-DC!): Siberia, "montis e Jablonsi Chrebet Dahuriæ", leg. Redowsky.

2n= (1) 64 (8x). - Europe (N), Siberia (N), Far East (N), Alaska, Canada, Greenland. - Numerous reports, more than 50 counts.
(2) 80 (10x). - Alaska, Canada, Greenland. - At least five reports. Petrovsky: There are counts of decaploids also from northeastern Asia.
There seems to be good evidence for two ploidy levels (8x and 10x) in this very widespread and polymorphic species.
Not included: Reports of 2n = 64 (8x) from Iceland (Löve and Löve 1956b; Löve 1970a). See notes. A report of 2n = ca. 75 from Alaska (Rollins 1966, for D. hirta) should be controlled by voucher.

Geography: Circumboreal-polar: NOR RUS SIB RFE ALA CAN GRL.

Notes: The name. - Russian authors have until recently almost uniformly applied the name Draba hirta L. to this species, whereas that name for long has been abandoned in northwestern Europe, Greenland, and North America. This critical name must either be rejected or supported by a type specimen. No type information is given in the protologue, i.e., there are no literary original sources. The known original plant material is Herb. Linn. 823.12 (LINN) with two plants. The left-hand plant is marked "Lapp." (Lapland, i.e., northern Sweden) and may possibly belong to D. glabella as currently understood but this will be difficult to decide due to the poor condition of the material. The right-hand plant is in even worse condition and probably belongs to D. rupestris. The Paris Linnaean Lapland herbarium (LAPP) should be checked for additional possible material. The protologue is not detailed enough to indicate any species ("D. fol. subhirsutis, scapo unifolio, siliculis obliquis, pedicellatis"). There is no indication that Linnaeus had the species currently considered as D. glabella, or predominantly this species, in mind when he described D. hirta. There are two options: Either to select the "Lapp." specimen as lectotype of D. hirta L. and in addition to select an epitype confirming this as a priority name for what more recently has been named as D. glabella or D. daurica (at least outside Russia), or to reject the name D. hirta L. Our principal opinion is that the name D. hirta has been applied and misapplied in too many meanings to have any current established usage outside Russia. It should therefore be proposed rejected. There are, however, two arguments against a rejection. (1) The IBC nomenclatural committee is notoriously reluctant to reject Linnaean species names. (2) Rejection of the name D. hirta may bring the name D. norvegica Gunnerus 1776, for a long time applied as the assumed priority name for another species, into the picture. This second oldest name most probably must be referred to D. glabella as currently considered. This will lead to significant instability in nomenclature. The simplest solution after all may be to reinstate D. hirta L. 1759 with an epitype, with D. norvegica Gunnerus 1776, D. glabella Pursh 1813, and D. daurica DC. 1821 as synonyms, and to reinstate the name D. rupestris W.T. Aiton 1812 for the other species.

The Siberian name D. daurica was applied regularly for the species in question in northern Europe until the 1990s (e.g., Elven 1994). The probable type of D. daurica corresponds well with North American D. glabella. The name D. glabella has priority if D. hirta is rejected and if D. norvegica can be avoided. We apply the name D. glabella for this species until these problems are solved.

Draba glabella in Iceland. - Draba glabella (or D. daurica) has regularly been reported from Iceland (e.g., Löve 1970a; Hultén and Fries 1986; Walters and Akeroyd 1993) but Jalas et al. (1996) discounted Icelandic presence. Numerous Icelandic collections have been annotated as this species (AMNH, C, ICEL, O) but with two exceptions they are misidentified. The exceptions are rather interesting. They are two specimens (ICEL) of very characteristic D. glabella stated to have been collected by Á. Löve in the 1930s in a little accessible locality in the (arctic) northwestern peninsulas of Iceland. This part of Iceland is one of those where Löve assumed glacial refugia. The specimens differ markedly from all other Icelandic Draba. The species has been searched for but not refound in the locality. The area is the same as for other very dubious Löve records from the same year (e.g., Eriophorum russeolum and Poa arctica). He made and vouchered several other such 'sensational' finds of close to ten species, all of them never refound there or elsewhere in Iceland. The evidence points towards these specimens having been collected somewhere else, probably in central or northern Scandinavia. This is either a case of massive confusion of material or a case of conscious 'planting' of specimens. The reports of octoploid chromosome numbers for D. glabella from Iceland (see above) support the suspicion that the records are conscious fakes. There is no known octoploid white-flowered Draba species reliably documented from Iceland.