Panarctic Flora

342001-03 The Calamagrostis purpurascens aggregate C. arctica, C. purpurascens, C. sesquiflora

Geography: Asian (NE) - amphi-Beringian - North American - amphi-Atlantic (W).

Notes: Elven and Murray: The Calamagrostis purpurascens aggregate is a widely distributed group of presumed sexual tetraploids and presumed agamospermous higher polyploids (6x, 8x, perhaps 12x). Tzvelev (PAF proposal) accepted five species: C. arctica, C. maltei, C. poluninii, C. purpurascens, and C. sesquiflora. Three of these are reported from northern and northeastern Asia, all five from North America including Greenland. For North America, Greene (1980, 1984) accepted one species with two varieties: var. purpurascens (including C. maltei) and var. laricina (including C. poluninii). However, he did not consider the Beringian plants (C. arctica, C. sesquiflora) in his study. Soreng and Greene in Soreng et al. (2003) accepted Greene's races and in addition C. sesquiflora in which they included C. arctica. Marr et al. (2007) did not accept var. laricina but commented on C. arctica. Whether the high polyploids are autoploids, alloploids, or a combination is unknown, but some taxa are suggested from morphological evidence to be alloploids.

Tetraploids are known from northeastern Asia (C. sesquiflora and within C. purpurascens s. lat.; several sources), northwestern Alaska (within C. purpurascens s. lat.; Johnson and Packer 1968), the Queen Charlotte Islands in western Canada (Taylor and Mulligan 1968 as subsp. tasuensis Calder & R.L. Taylor, included in C. sesquiflora by Soreng and Greene in Soreng et al. 2003, Marr et al. 2007, and by us), and possibly from eastern North America (within C. purpurascens s. lat.; Greene 1980, 1984, one count). An unambiguous delimitation of the tetraploids, except for C. sesquiflora (including subsp. tasuensis), is not yet possible. Only pollen-sterile and high-polyploid plants are known from arctic Canada, making it nearly certain that the type of C. purpurascens belongs in the agamospermous part of the aggregate. Pollen-fertile and therefore presumably tetraploid plants of C. purpurascens from northwestern North America do not differ in other morphological characters from pollen-sterile ones (many specimens at ALA inspected) and autopolyploidy is suspected. Calamagrostis purpurascens s. str. should therefore include, by our criteria and until further studied, both sexual tetraploids and agamospermous higher polyploids.

We follow Tzvelev, Soreng and Greene in Soreng et al., and Marr et al. in their acceptance of C. sesquiflora. We accept the viewpoint of Greene and of Soreng and Greene that there are two geographically major races within North American C. purpurascens but we treat them as parapatric subspecies. We deviate from Soreng and Greene by accepting the morphologically distinct and probably high-polyploid and agamospermic C. arctica apart from C. sesquiflora.