670904 Braya glabella Richardson
- Richardson in Franklin, Narr. Journey Polar Sea (Bot. App.): 743 (1823). Holotype (BM): Canada: Nunavut, Copper Mountains, leg. Richardson.
Geography: Circumpolar - Cordilleran.
Notes: Elven and Murray: Harris (1985) synonymized Braya purpurascens, B. henryae, B. bartlettiana, and some others with B. glabella. This treatment was adopted by Rollins (1993) and found support in the molecular data and the phylogeny presented by Warwick et al. (2003b).
The proposed taxa of B. glabella s. lat. have rather distinctive geographical patterns. The most widespread one is the high-arctic circumpolar B. purpurascens, the next most widespread one is the low-arctic and boreal North American B. glabella s. str., whereas three others are fairly local: B. henryae in the northernmost Cordilleras (morphologically close to B. purpurascens), B. bartlettiana in American Beringia, and B. aenea subsp. pseudoaenea in Asian Beringia (both close to B. glabella s. str.). Tetraploids (2n = 28) are known from Beringia, octoploids (2n = 56) from there and everywhere else, and dodecaploid (2n = 84) from Siberian Putorana. A late Pleistocene to early postglacial differentiation based on refugial situations is a hypothesis for the (slight) differentiation. A late divergence could also explain the lack of differentiation in the gene sequences analysed yet. At the same time it could support recognition of subspecies (i.e., regional races) rather than species.
We primarily would prefer treatment as two species: B. purpurascens (including B. henryae) and B. glabella (including B. bartlettiana and B. aenea subsp. pseudoaenea). The key, we believe, is whether one has seen them in the field and best of all at the same locality as is possible with B. glabella and B. purpurascens at Prudhoe Bay, Alaska, and in some northwestern Canadian localities. Mulligan commented that in his opinion, B. purpurascens should be synonymized with B. glabella as done by Harris, Rollins, and Warwick et al. We disagree, based on field experiences with both plants from several regions where they keep distinct even when sympatric or nearly sympatric, and on a survey of herbarium material. The high-arctic B. purpurascens retains its characters when grown under more moderate conditions similar to those experienced by B. glabella. We have not seen transitions documented. Even if the differential characters are few and mainly quantitative, the major parts of the material we have inspected in main collections from all parts of the range can relatively easily be divided into two parts with different but overlapping geographical patterns. For the Checklist, we accept Cody's (1994) recombination of B. purpurascens as a subspecies. This was done also by Harris (2010) for Flora of North America.