Panarctic Flora

500722 Saxifraga rosacea subsp. rosacea Moench

• Moench, Methodus: 106 (1794). Neotype (K): Germany. Ex Herb. Gay. "Saxifraga decipiens Ehrh. Rauhe Alb in Würtenburg. Endres misit, Aug. 1828." (Webb and Gornall 1989: 183).

2n= 32-64 (4x-8x). - Europe. - Several reports.
Löve and Löve (1975a) reported this species to be uniformly octoploid (2n = 64) and they listed only reports supporting this.

Geography: European (W): ICE.

Notes: Zhmylev: Saxifraga rosacea is a North Atlantic European, boreal species with a discontinuous distribution that includes the Faeroes, Iceland, Ireland, and the western parts of Central Europe. This species is one of the taxonomically most difficult of section Stoloniferae, which races are now being regarded as morphologically poorly delimited subspecies. Outwardly, S. rosacea looks like a lax form of S. cespitosa but differs from it by absence of glandular hairs on the leaves and pure white petals. The exception is S. rosacea subsp. hartii with glandular hairs on the leaves. This subspecies is reported to reach southern Greenland (Webb and Gornall 1989). However, the notes and counts of Löve and Löve (1951) should probably be referred to a lax form of S. cespitosa.

Elven and Solstad: Saxifraga rosacea subsp. hartii (D.A. Webb) D.A. Webb (S. hartii D.A. Webb) is considered by Webb to be a local race in northwestern Ireland and probably not very relevant for the arctic regions (e.g., Greenland). As for Greenland, Brouillet and Elvander (2009b) concluded that: "Reports of S. rosacea Moench from southern Greenland are misidentifications of this species" i.e., [S. cespitosa].

The question of S. rosacea and its delimitation from S. cespitosa remains unsolved and concerns especially the "laxiuscula" plants. The Löves assigned all decaploids to S. cespitosa (incl. the "laxiuscula" entity) and octoploids to S. rosacea. They reported both S. rosacea and the "laxiuscula" entity from Iceland but only the "laxiuscula" entity from Greenland (where only decaploids are known). Kristinsson (2008) accepted neither entity from Iceland but divided the material on S. cespitosa and S. hypnoides. The comparatively low-ploid Icelandic counts listed by the Löves might stem from S. hypnoides. However, Jalas et al. (1999) mapped S. rosacea from Iceland, including the arctic parts. We have seen numerous Icelandic plants, in the field and herbaria, that conform to S. rosacea as described and that differ from both S. cespitosa and S. hypnoides. All three species are common in Iceland and seem to keep distinct there.