Panarctic Flora

342002 Calamagrostis arctica Vasey

• Vasey, U.S.D.A. Div. Bot. Bull. 13, 2: t. 55 (1893). Holotype (US): Alaska: the Pribilof Islands, St. Paul Island, 31. Aug. 1891, leg. J.M. Macoun 38. - Calamagrostis purpurascens subsp. arctica (Vasey) Hultén, Acta Univ. Lund., n. s., sect. 2, 38, 1: 170 (1942).

2n= 56 (8x). - Far East (East Chukotka). - Zhukova (1969).

Geography: Amphi-Beringian: RFE ALA.

Notes: Elven, Murray, Tzvelev, and Yurtsev: The otherwise Chukotkan Calamagrostis arctica is now documented from four localities on the Seward Peninsula in western Alaska (ALA) and has probably been overlooked. As the octoploid chromosome number is based on only one count, it is uncertain what ploidy levels are found in C. arctica. We have observed only aborting anthers in the material and assume C. arctica to be agamospermous and probably constantly more high-ploid than tetraploid.

Tzvelev and Yurtsev both argued for C. arctica as a recognizable species and Yurtsev summarized the differences. Morphologically, it is more similar to C. sesquiflora and differs from C. purpurascens in leaves broader and scabrous (not short-pubescent) above, glumes subulate and red, and lemmas (in C. arctica as well as in C. sesquiflora) with long, almost subulate teeth, whereas in C. purpurascens the teeth are weakly expressed. In addition, C. arctica is suboceanic and occurs on acidic substrates, whereas C. purpurascens is continental and occurs mainly on calcareous substrates. They both occur in the same regions, e.g., on the Chukchi and Seward peninsulas, but never together due to the edaphic preferences.

Murray and Elven studied material (ALA) from Alaska and Chukotka and concluded that C. arctica disjunctly differs from C. purpurascens in at least three characters: leaf upper surface setose vs. soft-pubescent, glumes long pointed and acuminate vs. evenly tapering and acute, and glumes dark purple vs. pale pink or lilac. Two of these characters are intermediate between C. purpurascens and C. sesquiflora, and C. arctica also partly takes an intermediate position geographically. It may be interpreted as a high-ploid agamospermous derivative of a hybrid between tetraploid C. sesquiflora and tetraploid C. purpurascens. Yurtsev commented that the origin of C. arctica from C. sesquiflora cannot be excluded but that the second parent (in case of allopolyploidy) may by no means be C. purpurascens.

Neither of us accept the inclusion of C. arctica in C. sesquiflora as done by Soreng and Greene in Soreng et al. (2003) and Marr et al. (2007). We find C. arctica consistently different from both C. sesquiflora and C. purpurascens in several characters and without any signs of transitions among these three. We accept C. arctica as species inasmuch as it occurs entirely within the range of C. purpurascens and keeps constantly distinct.