Panarctic Flora

421213 Silene uralensis (Rupr.) Bocquet

• Bocquet, Candollea 22: 26 (1967). - Gastrolychnis uralensis Rupr., Verbr. Pfl. Ural: 30 (1850). Holotype (LE!): European Russia: "Westfusse des Ural im 67 1/2 in der Nähe des Flusses Porotschjadyr", 23-24. July 1848, leg. Ruprecht. Illustrated in Ruprecht, Fl. Bor.-Ural.: t. 1, f. 2. 1854. - Gastrolychnis apetala subsp. uralensis (Rupr.) Á. Löve & D. Löve, Bot. Not. 128: 510 (1976).

Geography: Circumpolar-alpine.

Notes: Elven, Murray, and Petrovsky: Tzvelev (2000b) recognized the diploid parts of Silene uralensis s. lat. as a series of species, whereas the majority of authors have considered these plants to belong within one polymorphic species with two or more races. Compared with Bocquet (1967) and Morton (2005c), we exclude subsp. apetala (= S. wahlbergella, see above) and the tetraploid subsp. porsildii (= S. soczavana, see below).

At least five epithets are relevant in the diploid part of the complex: (1) "uniflora" Ledeb. 1815 for a species in Lychnis, described from Transbaicalia in southeastern Siberia; (2) "uralensis" Rupr. 1850 for a species in Gastrolychnis, described from the northern Urals; (3) "arctica" Th. Fr. 1870 for a variety in Wahlbergella, Hultén 1944 for a subspecies in Melandrium, described from Svalbard in Norway; (4) "attenuata" Farr 1904 for species in Lychnis, described from northwestern North America; and (5) "violascens" Tolm. 1971 for a species in Gastrolychnis, described from Yakutia in Siberia. The earliest name - "uniflora" - is inapplicable at species level in Silene as it is predated by S. uniflora Roth 1794. The earliest name available within Silene for a collective species (irrespective of how collective), is S. uralensis as made clear by Bocquet (1967).

The infraspecific structure among the diploid plants is much more debatable. Kozhanchikov and Tolmachev (1971) accepted only one species - Gastrolychnis apetala - without subdivisions for the arctic parts of Russia. Hultén (1968a) named only one taxon - Melandrium apetalum subsp. arcticum - from Alaska and the Yukon Territory, mapped it as circumpolar, but accepted a Scandinavian subsp. apetalum (now S. wahlbergella). Böcher et al. (1978) accepted the same taxon and name for Greenland as did Porsild and Cody (1980) for the arctic parts of Canada. These authors accepted, however, a Cordilleran race: Melandrium apetalum subsp. attenuatum. Morton (2005c) accepted a S. uralensis subsp. uralensis including "arctica" and "attenuata" for North America. Chater et al. (1993) accepted S. uralensis subsp. uralensis (circumpolar) and subsp. apetala (= S. wahlbergella) for Europe. Kurtto (2001c) accepted S. uralensis (circumpolar) and S. wahlbergella for northwestern Europe. The monographer - Bocquet (1967, 1969) - accepted S. uralensis with three northern diploid races: subsp. uralensis as circumpolar but replaced by subsp. apetala in Scandinavia and by subsp. arctica in Svalbard. A molecular study of this complex is under way (A. Petri, B. Oxelman etc.) but results are not yet available. Waiting for these, we have to rely on variation in morphology and ploidy levels. Our conclusions, provisional as they must be, are as follows:

We disagree with Bocquet's view that subsp. arctica is a local Svalbard plant, whereas all the other arctic plants belong to subsp. uralensis. Subspecies arctica is characterized by the very strongly inflated calyx, especially in fruit when it becomes globular or even broader than long, petals much emerging from calyx, capsules short (up to 15 mm), and upper parts of stems distinctly and often predominantly with glandular hairs. We have compared Svalbard plants with plants from arctic Russia (LE: Petrovsky and Elven), Greenland (ALA, CAN, DAO, O: Elven), arctic Canada and Alaska (ALA, CAN, DAO: Murray and Elven), and in the field in several regions. We can find no differences among these high-arctic circumpolar plants and consider them one uniform taxon. The only epithet relevant for this taxon is "arctica".

Nearly all authors from Bocquet (1967) to Tzvelev (2000b), Kurtto (2001c), and Morton (2005c) have assumed the Urals plant (subsp. uralensis) and the more arctic one (our subsp. arctica) to be the same. Subspecies arctica differs from the plant of the Urals (subsp. uralensis, including the type which we have inspected), and also from more southern arctic and alpine plants in northeastern Asia and in North America, in the characters referred above. We have compared material (LE, O) from the zone of contact from the Urals through the Yugorski Peninsula and Vaigach to Novaya Zemlya. We see two morphologically different taxa. In this region, subsp. uralensis is the only one documented from mainland northeastern European Russia (Polar Ural and the Yugorski Peninsula) and northern Siberia (the Yamal-Gydan area and at least east to the Lena River). Subspecies arctica is the only one documented from Novaya Zemlya. On the island of Vaigach, between the Russian mainland and Novaya Zemlya, both taxa occur but remain distinct. No transition was found in the large material in LE.

Subspecies uralensis is characterized by the calyx being not strongly inflated and usually longer than broad even in fruit, capsules short (up to 15 mm), petals slightly emerging from the calyx (less so than in the high-arctic race), but stems hairy as in the high-arctic race. Tzvelev (2000b) suggested that the name "uniflora" has priority for this plant (but only within Gastrolychnis, not in Silene) and that the name G. uniflora (based on Lychnis uniflora) should replace G. uralensis. We have inspected the type of Lychnis uniflora and suspect that it may belong to a small-grown S. violascens (see below).

One question is how widely distributed is subsp. uralensis. We have compared northern North American and northern Eurasian plants and find them essentially similar. We therefore follow Morton (2005c) in extending subsp. uralensis across all of North America to southern and western Greenland. In Greenland, there is an overlap in the ranges of northern subsp. arctica and more southern subsp. uralensis at 70-71N but no obvious transitional plants. In the Canadian Arctic Archipelago, subsp. arctica is the most common race but both subspecies are present in Banks, Victoria, and Baffin islands, with subsp. uralensis in the southern parts and subsp. arctica in the northern ones. There is some overlap, e.g., subsp. arctica on the Hudson Strait Mill Island, but no obvious morphological transition. Our conclusion is that two taxa (at least as races) are justified in a circumpolar context. The only area where we have not found it equally easy to sort material is the northern parts of Alaska and the Yukon Territory.