Panarctic Flora

420116 Stellaria longipes Goldie

• Goldie, Edinburgh Philos. J. 6: 327 (1822). Neotype (E): Canada: Ontario, Kingston, near Odessa, on natural prairie on limestone, 13. June 1972, leg. J.K. Morton NA5101 (Chinnappa and Morton 1991: 131).

2n= 26-107 (2x-8x). - Europe: Svalbard (8x), Scandinavia (8x); Siberia (N, S; 4x, 8x); Far East (N; 3x, 4x, 6x, 8x); Alaska (6x, 7x, 8x); Canada: Yukon-Mackenzie (6x, 8x), Hudson Bay-Ungava (4x, 7x, 8x), Arctic Archipelago (4x, ca. 6x, 8x); non-arctic Canada/U.S.A. (2x, 4x, 6x, 8x); Greenland (4x, 6x, 7x, 8x). Also several reports of numbers between the ploidy levels.
Reports under the following names with indicated ploidy levels (exact or approximate) and rough numbers of reports: S. arenicola 5-6x (1); S. ciliatosepala 5-6x (5), 6x (4), 7x (3), 8x (5); S. crassipes 6-7x (1), 7x (1), 8x (8); S. davurica 6x (2); S. edwardsii 3x (1), 4x (1), 5-6x (1), 6x (2), 7x (3), 8x (4); S. laeta 4x (2), 5-6x (1), 6x (3), 6-7x (2), 7x (1); S. laxmannii 5x (1); S. longipes s. str. 2x (1), 4x (6), 6x (3), 8x (3); S. longipes s. lat. 4x (4), 5-6x (2), 6x (1), 7x (1), 8x (5); S. monantha 3x (1), 4x (6), 6x (2), 8x (7); S. peduncularis 5-6x (3), 6x (2), 8x (2); S. stricta 6x (1); S. subvestita 4x (1).
Some authors of chromosome number reports have assumed a strict base number of x = 13 and have only found exact or near exact multiples of this. Other authors have reported many intermediate (e.g., 2n = 72 and 84) or approximate numbers. Species names may have been assigned dependent on the chromosome number found, especially after some proposals by the Löves (e.g., S. crassipes as octoploid and S. laeta as hexaploid). Names have sometimes been applied to plants far from where they were described and are assumed to be present (at least two names for local American species attempted applied also in Asia). Accordingly, the species names in the chromosome number reports (and in other reports) do not mean much.
The only report of a diploid number is Cai and Chinnappa (1989), specified to be for S. longipes s. str. from Canada. Otherwise, tetraploids are known from Siberia, the Far East, non-Beringian North America, and Greenland, whereas only hexaploids to octoploids are known from northern Europe and Beringian America.

Geography: Circumboreal-polar: NOR RUS SIB RFE ALA CAN GRL.

Notes: Elven and Petrovsky: There are two approaches to this extraordinarily polymorphic, complicated polyploid, and circumpolar complex:

(a) Hultén (1943a) followed many previous authors and accepted several parapatric to sympatric species. Porsild (1963) detailed this treatment for North America; Böcher (1951b), Philipp (1972), and Böcher et al. (1978) for Greenland; and Matzenko (1971), Antonova and Petrovsky (1986), and Vlasova (1993) for the Russian areas. The diagnostic differences reported by these authors are mainly in the absence or presence of pubescence on stems and sometimes leaves, whether the plants are glaucous or not, leaf shape, whether the bracts in the inflorescence are hyaline or hyaline-margined or not, numbers of flowers in the inflorescence, peduncle length, sepal shape, degree and location of pubescence on sepals, and seed surface features. Such characters are considered diagnostic among species elsewhere in Stellaria. This species aggregate approach was until the 1990s applied in the major floristic surveys for all regions (Hultén 1968a for Alaska and the Yukon Territory; Porsild 1957 and Porsild and Cody 1980 for Canada; Böcher et al. 1978 for Greenland; Chater and Heywood 1993 for Europe; Matzenko 1971 for arctic Russia; Vlasova 1993 for Siberia; Pavlova and Bezdeleva 1996 for the Russian Far East; Hultén and Fries 1986 for the entire circumpolar range). Some duplicate naming among regions is suspected.

Chater and Heywood (1993) indirectly referred to the other approach (below) in the following comment: "The three species i.e., [S. longipes, S. crassipes, and S. ciliatosepala as accepted from Europe] ... are morphologically quite distinct in Europe, but not in North America, and a satisfactory interpretation of the group cannot be made without further investigation throughout the range".

(b) From the 1970s onwards, Chinnappa, Morton, and collaborators have studied morphology, cytology, reproduction, and isoenzyme variation in the complex, especially in North America (e.g., Chinnappa and Morton 1974, 1976, 1984, 1991; Chinnappa 1985; MacDonald and Chinnappa 1988). These authors have found (according to the summary by Borgen and Often 2001): "great phenotypic plasticity, low correlation between cytotype and morphology, and high fertility in crosses? among cytotypes ... It seems therefore appropriate to treat the complex as one polymorphic species". They assumed the complex to be a polyploid reticulum based on several parental diploids, S. longifolia probably one of them (see Cai et al. 1990), and diploid S. longipes must be another. This approach has been followed in later floristic surveys (e.g., Borgen and Often 2001 for northwestern Europe; Morton 2005b for North America; Brysting et al. 2007 for the Canadian Arctic Archipelago).

The Chinnappa-Morton interpretation of the genetic situation is well supported by investigations. Still, field botanists in the Arctic tend to see structured morphological variation, to find it possible to recognize several taxa, to assign the majority of the plants they find to them, and to name them. This contrast between the field botanists' impression of more or less (mostly less) distinct taxa and the experimental botanists' analysis of continuous variation could perhaps be solved by experimental work with several molecular markers on carefully sampled and morphologically analysed and sorted material. We have decided to accept only the collective species, but to enter the other named species provisionally as subspecific "taxa" but assuming that if taxa are recognizable, it is probably as complex allopolyploid species rather than as, e.g., subspecies. We have merged in a few cases where synonymization between Asian and North American names seems reasonable. Indications of chromosome numbers and ploidy levels are based on the names given in the different sources and are probably thoroughly unreliable.