Panarctic Flora

500205 Chrysosplenium rosendahlii Packer

• Packer, Canad. J. Bot. 41: 89 (1963). Holotype (DAO): Canada: Nunavut, Somerset Island, N. shore of Hazard Inlet, 7203'N, 9407'W, 24. July 1958, leg. D.B.O. Savile 3619.

2n= 96 (16x, x = 6). - Far East (Wrangel Island), Canada. - Packer (1963); Zhukova et al. (1973); Petrovsky and Zhukova (1981).

Geography: Amphi-Beringian - North American (N): RFE ALA CAN.

Notes: Aiken, Murray, and Elven: Chrysosplenium rosendahlii is enigmatic and has had a peculiar fate in North American literature since Packer's (1963) publication. Löve and Löve (1975a) accepted it as a species, naturally from their principles as it differs in ploidy level. Scoggan (1978c) reduced it to a synonym of his very widely circumscribed C. alternifolium together with, e.g., C. tetrandrum. Porsild and Cody (1980) excluded it altogether, even from the synonymy. The majority of authors from then on have done the same until it was reinstated as an accepted species in Flora of North America (Freeman and Levsen 2009).

We have looked at all the specimens annotated by Packer in ALA, CAN, and DAO. The main characters separating C. rosendahlii from C. tetrandrum are, as stated by Packer (1963), the occurrence of more than four stamens (mostly eight) and the unequal size of sepals. The latter character we found difficult to confirm on many of the herbarium specimens. The former character is distinct but there is variation between four and eight stamens even in the type specimen. Freeman and Levsen (2009) also noted larger seeds, 0.8-1.1 mm vs. 0.5-0.8 mm.In Canada and Alaska, the range of C. rosendahlii is enclosed within that of the far more common C. tetrandrum, from northern Alaska and western Banks Island east to western Baffin Island, north to Somerset Island, and south to Victoria Island, Melville Peninsula, and Southampton Island. The main argument for retention of C. rosendahlii is its high ploidy level (see also Packer 1959). Throughout its wide range, C. tetrandrum has been documented only with 2n = 24. Chrysosplenium rosendahlii may well be an allopolyploid with C. tetrandrum in its parentage.

Occurrence of C. rosendahlii in northern Alaska is confirmed as specimens from two localities (Prudhoe Bay and near Beaufort Lagoon, ALA) have been identified as such by Packer and are accepted by us.

Petrovsky: The Wrangel Island material is distinctive although close to C. alternifolium subsp. arctomontanum morphologically. Perhaps there is a better case for C. rosendahlii on Wrangel Island than in Canada?

Murray: I think it is a good taxon and undeniably in Alaskan and Canadian Arctic.