Panarctic Flora

630504 Astragalus alpinus L.

• L., Sp. Pl.: 760 (1753). Lectotype: Scheuchzer, Uredif. Helv.: f. 7. 1723; epitype (BM): Switzerland: "Joch-Alpe prope Churwalden", July 1863, leg. Brügger (Podlech in Turland and Jarvis 1997: 463).

2n= Rough number of reports indicated for each region:
(1) 16 (2x). - Western Europe (N-4, C/S-5), European Russia (N-3), Siberia (C/S-5), Far East (N-4), Canada (N/C-1), U.S.A. (NW-1). Reported for "alpinus" and "subpolaris".
(2) 32 (4x). - Siberia (N-3), Far East (N-11), Alaska (4), Canada (NW-2, N-3). Reported for "alaskanus", "alpinus", and "arcticus".
See notes.

Geography: Circumpolar-alpine.

Notes: Elven and Murray: Astragalus alpinus is the most widespread species of the genus and in some arctic regions the only native species of Fabaceae. Some polymorphy is to be expected. Yurtsev proposed three races (subspecies) for inclusion and reported the following ploidy levels and geographical ranges: a tetraploid subsp. alpinus in northeastern Asia (Siberia and the Far East) and North America; a diploid and tetraploid subsp. arcticus in northern Europe, northern Asia, North America, and northeastern Greenland; and a diploid subsp. alaskanus in South and East Chukotka and in Alaska. Yurtsev (1986: 39) mapped the ranges of these in northern Russia. There are several problems with Yurtsev's proposal.

(1) Subspecies alpinus is almost certainly diploid, not tetraploid as stated by Yurtsev. Astragalus alpinus was described from mountains in northern Scandinavia and Switzerland (Linnaeus 1753: "Alpibus Lapponicis, Helveticis"), now based on lecto and epitypes from Swiss plants (Podlech in Turland and Jarvis 1997). Only the diploid number has been found in European plants. When Yurtsev concluded that subsp. alpinus is tetraploid he must have had another plant in mind than the A. alpinus as covered by the protologue and type.

(2) The name "arcticus" is based on another nomenclatural type than assumed by Yurtsev. Variety arcticus (Sondén) Lindman 1918 is based on f. arcticus Sondén 1907, described from the Torneträsk area in Torne lappmark, northern Sweden. Subspecies arcticus (Bunge) Lindman 1926, and Hultén's similar, superfluous combination from 1947, are based on Astragalus arcticus Bunge 1869, described from northeastern Fennoscandia, northern European Russia, and northeastern Siberia. The names f./var arcticus and subsp. arcticus are thereby based on different types, making the name subsp. arcticus an illegitime homonym according to the Code (ICBN article 53.4, McNeill et al. 2006). The name var. arcticus is applicable for the northern Fennoscandian plant and its meaning is clear. However, the two meanings of the name "arcticus" probably refer to the same taxon, and Yurtsev may be right in excluding subsp. alpinus from arctic Europe and in reporting var. arcticus (as subsp. arcticus) as widespread. Subspecies alpinus occurs in central and western Europe and reaches far northwards in coastal parts of Norway but not the Arctic. It is replaced in the major parts of northern and northeastern Europe by var. arcticus.

(3) Subspecies alaskanus is probably entirely or mainly tetraploid, not diploid as stated by Yurtsev. The six reports from within the range originally reported for subsp. alaskanus, i.e., Alaska and the Yukon Territory, are of tetraploids (see above).

The three proposed races differ morphologically but in a complicated pattern. The best diagnostic difference between var. arcticus and subsp. alpinus is in petal proportions: keel shorter than the wings and much shorter than the banner vs. longer than the wings and only slightly shorter than the banner. This may seem a small difference but it makes the flowers and flower heads different at a glance; the protruding banners and hidden keels of var. arcticus are prominent features. The flowers of var. arcticus are mostly purple or deep blue vs. blue to pale blue or nearly white in subsp. alpinus. Other reported diagnostic differences (see, e.g., Chater 1968; Elven et al. 2005) are not fully reliable: calyx teeth triangular and much shorter than the tubular part of the calyx in var. arcticus vs. almost linear and about as long as the tubular part of the calyx in subsp. alpinus, and some difference in growth form (var. arcticus is more compact) and in leaf colour (var. arcticus is darker green). There is a comparatively narrow belt with transitional forms between these two races in west-central Fennoscandia.

Subspecies alaskanus differs from subsp. alpinus in the same characters: the keel distinctly shorter than the banner (by 2 mm or more) and mostly also shorter than the wings. It differs from both subsp. alpinus and var. arcticus in narrower wings (ca. 2 mm broad vs. more than 3 mm in the others). Also this difference is pronounced as the wings hide the keel in subsp. alpinus and var. arcticus but not in subsp. alaskanus. In addition, subsp. alaskanus differs in an extended raceme in plants growing under benign conditions, but short, compact racemes are common in alpine plants of subsp. alaskanus. The flowers are blue to pale blue, not purple as in the majority of var. arcticus.

We have confirmed plants with the morphological pattern of subsp. alpinus from central and western Europe north to the southern parts of northern Norway, from northeastern Asia, and from throughout northern North America and the western Cordilleras. There is probably a connection through European Russia and Siberia but we have not had material available for inspection. The morphological pattern of var. arcticus is found in plants from northern and eastern Fennoscandia, northern European Russia, northern Siberia and Russian Far East, and northern Alaska, mainly outside the range of subsp. alpinus but overlapping with it at least in Fennoscandia, Siberia, and Alaska. The morphological pattern of subsp. alaskanus has only been found in plants from the Yukon Territory, Alaska, and perhaps the Chukchi Peninsula, in Alaska and the Yukon mainly from the western and central parts, overlapping with the ranges of subsp. alpinus and var. arcticus in the north and replaced by subsp. alpinus southwards from the southern Yukon Territory. The material that closely conforms to the characters specified for subsp. alaskanus by Hultén (1947, 1968a) is in a minority even in Alaska, compared with material with more diffuse characteristics. The ambiguities concerning Russian material are marked in the distribution table.

The two ploidy levels differ in their geographical distribution but not in good accordance with the proposed races based on morphology, neither in our concept nor in Yurtsev's. Only diploids are known from Europe (from both subsp. alpinus and var. arcticus), central and southern Siberia (subsp. alpinus), and from non-arctic Cordilleran North America (possibly subsp. alpinus). Only tetraploids are known from northern Siberia (mainly var. arcticus), Alaska (subsp. alaskanus and var. arcticus), and arctic Canada (probably subsp. alpinus). Both diploids and tetraploids are known from the Russian Far East. Löve and Löve (1976a) separated the two ploidy levels on two species: the diploids on A. alpinus s. str., the tetraploids on A. astragalinus, the latter name assumed to be a synonym for subsp. alaskanus. Their solution is not supported by morphology.

Welsh (2007: 48-49) accepted two varieties in A. alpinus from North America, var. alpinus (including both var. arcticus and subsp. alaskanus) and a non-arctic eastern var. brunetianus Fernald.