670903 Braya humilis (C.A. Mey.) B.L. Rob.
- B.L. Rob. in A. Gray, Syn. Fl. N. Amer. 1, 1: 141 (1895). - Sisymbrium humile C.A. Mey. in Ledeb., Icon. Pl. 2: 16, t. 147 (1830). Holotype (LE): Siberia: the Altai, "in argillosis subsalsis ad fluvios Kenlyk, Kan, Jebagan; ad fluvium Tschuja", 1826, leg. Bunge 1033. - Torularia humilis (C.A. Mey.) O.E. Schulz, Repert. Spec. Nov. Regni Veg. Beih. 12: 390 (1922). - Neotorularia humilis (C.A. Mey.) Hedge & J. Léonard, Bull. Jard. Bot. Belg. 56: 394 (1986).
2n=
28 (4x). - Canada (NWT, Alberta, British Columbia). - Mulligan (1965a, 2002); Mulligan et al. (1972).
The counts are reported only for "humilis" in the original publications and by Mulligan (2002). They may belong to one or both of the two non-arctic tetraploid subspecies accepted by Harris (2006), see below.
Geography: Asian (N/C) - amphi-Beringian - North American (N).
Notes: Several species or subspecies have been proposed to occur in the arctic regions (Böcher 1950, 1956, 1966a; Hultén 1971a; Harris 2006) within what Harris (1985) and Rollins (1993) treated collectively as Braya humilis. Warwick et al. (2003b) discussed the variation within the species. They mentioned that tetraploids (2n = 28) are restricted to non-glaciated (Beringian) parts of Alaska and northwestern Canada. The tetraploids seem to be large-flowered outcrossers. Harris (2006) described these as two subspecies: subsp. maccallae J.G. Harris and subsp. porsildii J.G. Harris. Hexaploids (2n = 42) and octoploids (2n = 56) are widespread. These are variable and seem to be small-flowered inbreeders. Harris considered them collectively as subsp. humilis, except that he described a subsp. ellesmerensis J.G. Harris from Ellesmere Island. Decaploids (2n = 70) are restricted, as far as known, to northern Greenland, show low fertility, and were included in subsp. humilis by Harris. The appreciable variation in morphology and ploidy levels may be due to comparatively recent hybridization and polyploidization, perhaps only within B. humilis s. lat.
Böcher (1966a, 1973) was of another opinion and assumed taxonomic allopolyploidy. A convincing geographical structure of the morphological variation remains to be proved. The combination of differences in morphological and reproductive traits mentioned by Warwick et al. (2003b) and Harris (2006) should be more closely studied for each of the ploidy levels (4x, 6x, 8x, 10x) and compared with the range differences. If (taxonomic) autopolyploidy is involved, this should be documented by cytological studies. Merging of several ploidy levels within a single subspecies (subsp. humilis) should be argued for by documented gene flow among ploidy levels. There seems to have been a fairly recent (probably late Pleistocene) diversification within the species, not well reflected in the sequences studied yet. This group may illustrate a possible discrepancy between "old" taxonomy based predominantly on morphology, reproductive traits, and cytology (ploidy levels indicating reproductive barriers), and "new" taxonomy based predominantly on sequences and other molecular markers.
Besides the two subspecies accepted by Harris (2006) from the arctic parts (a and f below), the previously proposed taxa are entered provisionally (after subsp. humilis) as we are not convinced that they are without any reality.