Astragalus alpinus var. alpinus

Phaca astragalina fma. occidentalis (western) Gand. in Bull. Soc. Bot. Fr. 48. xviii. 1902. —"Colorado, ad Como, Park Co., alt. 9800 ped. (C. Crandall). Wyoming, ad Green Top (Nelson exs. no. 3263); Montana ad Yellowstone Park, alt. 9000 ped. (Rydberg); Idaho (Heller); New Mexico in alpinis (Wooton)."—Lectotypus, collected at Como, Park Co., Colorado, July 23, 1897, Crandall, LY!—Spms. from Como dated August, 1895, Crandall 896, NY, US, WIS, are possibly isotypi.

Astragalus alpinus fma. arctica [sic] Sond. in Svensk. Bot. Tidsk. 1 (1907): 233. 1908. —No locality given; no typus examined.

Astragalus andinus Jones, Rev. Astrag. 137. 1923, the name taken from Phaca andina (of the Rocky Mountains or northern Andes) Nutt. in herb. & ex T. & G., Fl. N. Amer. 1:345. 1838, in syn.—"Alpine summits of the Rocky Mountains, near the confines of perpetual snow (Thornburg’s Pass), about latitude 43°, Nuttall."—Lectotypus, labeled by Nuttall "Phaca *andina. In the perpetual snow line of the R. Mts. sources of the Shoshone., NY! isotypus, from "Thornburg’s Pass," BM!

Astragalus alpinus var. parvulus (smallish) Rouss. in Contrib. Lab. Bot. Univ. Montreal 24: 23. 1933.—"Bowdoin harbor, Baffin Land, June 10, 1922, MacMillan Expedition. Ralph Robinson 15 (type: Gray herbarium)."—Holotypus not found at GH in 1960.—A. alpinus fma. parvulus (Rouss.) Rouss. in Le Nat. Canad. 69 : 238. 1942.

Astragalus alpinus subsp. alaskanus (of Alaska) Hult., Fl. Alaska & Yukon 1082. 1944. —" ... specimens from the interior and Seward Peninsula ... "—No typus designated, but 15 collections at NY, US, cited by Hulten, were examined.—A. alpinus var. alaskanus (Hult.) Lepage in Amer. Midl. Nat. 46: 757. 1951.

Astragalus alpinus fma. Lepageanus (Rev. Ernest Lepage, 1905- ) Rouss. ap. Lepage in Le Nat. Canad. 77: 230. 1950.—"Quebec: Rivieres aux Melezes (57° 41' N., 69° 32' W.), plage sablonneuse, 11 août 1945, Dutilly & Lepage 14631 (TYPE, Herbier Marie-Victorin, Univ. Montreal)."—Holotypus not examined; described as a minor mutation with glabrous calyx and pod.

Astragalus alpinus var. alaskanus fma. albovestitus (white-clad) Lepage in Amer. Midi. Nat. 46: 758. 1951.—"Bering Strait Distr.: Kotzebue, Lepage 25410, Aug. 11, 1949."—Holotypus LCU!

The Austrian botanist Nicholas Jacquin (Collectanea 1: 323. 1786) neatly described the alpine milk-vetch as "gratissimus semper, et quia natura tenellus et quia adspectu decorus"; it is indeed an ever delightful, pretty but unassuming plant which speaks to the plant geographer of high, cool, and northern places. Wherever one may come upon it in its extensive range over arctic tundra and down through the mountains of the Northern Hemisphere, he will recognize the alpine milk-vetch by its slender, freely branching subterranean caudex, connate stipules, nodding flowers of characteristic shape, and by the pendulous, stipitate, trigonous but dorsally grooved pod pubescent nearly always with mixed black and white hairs. Typical var. alpinus is so well and widely known that it has seemed superfluous to cite representative material which may be found in plenty in all the larger herbaria of Europe and North America. In America the alpine milk-vetch is often sympatric and sometimes confused with forms of A. Robbinsii, a species similar in coloring and choice of habitat but easily distinguished by its superficial and determinate root-crown at the level of the soil, large, veiny, lower stipules, and a pod which is sometimes trigonous but only flattened and not sulcate dorsally. Care must also be taken to discriminate between A. alpinus and two dwarf species of Oxytropis, in Europe O. lapponica (Wahlenb.) Gay and in America O. deflexa var. foliolosa (Hook.) Barneby, which are not only similar in superficial appearance but not uncommonly found growing with it. The beaked keel, of course, distinguishes Oxytropis as a genus, and O. deflexa differs further by its free stipules and pod semibilocular by intrusion of a funicular flange and not by a true septum.

The alpine milk-vetch, in spite of its immense range, is not a highly variable species. The foliage is commonly green and thinly pubescent, but may be canescently hirsutulous, a phase known from scattered stations in America from Colorado to the Yukon Valley. The variation in stature and in the relative proportions of the petals has been analyzed from the point of view of the Alaskan flora by Hulten, who proposed for the relatively rank-growing, perhaps polyploid plant of the far Northwest (and probably of the Rocky Mountains) a subsp. alaskanus differing from subsp. alpinus in its elongate raceme of flowers with keel shorter than rather than equaling the banner and with narrow wings. However even in the Alps of Europe the wings of A. alpinus may vary from 1.4 to 3.4 mm. in width, with a variation of over 1 mm. recorded from plants mounted as parts of one collection, and this feature clearly has no systematic significance of itself. In Europe also the banner varies from at least 0.5 mm. shorter to 1 mm. longer than the keel. In the Rocky Mountains an elongate raceme, common in America but found only rarely in Europe, is not correlated with a keel short in proportion to the banner, nor with narrow wings. Not only are there short and long (more or less than 4 cm.) racemes in the same population, but the wings vary in the longer type from 0.8-3.2 mm. in width and the keel varies in the same group of plants from 0-2 mm. shorter than the banner. In North America, northward along the Cordillera from western Wyoming, A. alpinus is common in the timber belt and quite variable in stature and length of the racemes. Plants luxuriating in moist soil along brooks overhung with thickets of willow and shrubby cinquefoil, or on gravel bars of mountains torrents, as also in drier soils of lodgepole pine or aspen forest, tend to be drawn out, broad-leaved, and loosely flowered. But wherever a population of this sort, which corresponds closely to subsp. alaskanus Hult., passes directly and continuously out of moist shade onto a dry turfy bank or sunny exposure, the more compact growth-habit generally associated with the European plant is immediately regained. The subsp. alaskanus seems to be based on an arbitrarily selected series of relatively tall and lush specimens which even in Alaska and Yukon (cf. Hulten, 1944, maps 813, 813a) lack the geographic segregation which is normally expected of a subspecies.

A second variant of A. alpinus, possibly of greater racial importance since it is confined to the highest latitudes around the pole, has been described as A. arcticus Bge. It was maintained by Gontscharov as a distinct species under the epithet subpolaris, and recognized by Hulten (1944, 1. c.) as a subspecies coordinate with subsp. alaskanus. Bunge distinguished it from A. alpinus because of its regularly graduated, more richly colored petals, supposedly correlated with a more robust habit of growth. Gontscharov (op. cit. 40, in clave) brought out a further contrast between a fully bilocular pod in A. alpinus and a semibilocular one in A. subpolaris. About forty collections (over thirty from America) have been seen which would be referred to A. subpolaris because the wings surpass the keel in length. By no means all of these have bright purple flowers, and at least in America a flower of this type is not correlated with a relatively narrow septum. In Alaska the plants most closely resembling A. arcticus of Novaya Zemlya and northern Lapland are found chiefly along the north coast eastward from Behring Strait, but technically inseparable forms occur in the interior up to 3500 feet elevation. Hulten’s map (1944, no. 813b) shows correctly that the subsp. arcticus and alpinus are sympatric even at their northernmost limits, but leaves out the many puzzling specimens which fail to satisfy the criteria of proportions and coloring of the petals and cannot be considered typical of either subspecies. Such plants as Coville & Kearney 744 (US) with ample, brightly colored flowers but wings shorter than the keel, or A. E. Miller in 1919 (US) with small, pale flowers and wings longer than the keel, are real obstacles to recognition of distinct subspecies in Alaska. Hulten (1. c.) cites subsp. arcticus as extending east to Baffin and Labrador, but I have seen no material of the extreme type from east of Hudson Strait (A. P. Low 22,863, NY). Specimens near A. arcticus as ideally conceived have been collected as far south as Thunder Bay District, Ontario (Taylor & al. 679, GH), and these are particularly suggestive of an individual, independent variant not genetically linked to the arctic type. The differential characters are quite feebly correlated, and it seems preferable to regard A. arcticus as a minor variant. It should be added that the names applied to A. arcticus Bge. are all based on Phaca lapponica sensu DC., non Wahlenberg, of which a lectotypus has been designated above. Bunge’s interpretation of DeCandolle’s intention is here followed, but it should be remarked that DeCandolle cited Table 51 of Oeder’s Flora Danica as illustrating his species. The figure clearly shows wing-petals surpassed by a broad and prominent keel and seems to be a superlatively accurate portrait of A. alpinus sens. str.

Although A. alpinus seems to be most variable, at least in America, north of about lat. 60° N., there is some evidence also of incipient or possibly reliquial racial differentiation in the Rocky Mountains. In this region the relative proportons of the flower parts remain rather constant, but the whole flower varies considerably in size. With very few exceptions the populations found between the Medicine Bow Range in southern Wyoming and the Sangre de Cristo massif in northern New Mexico are perceptibly smaller-flowered than those occurring northward from the Wind River Mountains. It seems probable that the saddle in the Cordillera corresponding with the Red Desert has been as significant a barrier to A. alpinus as to many other plant species. The following figures (in mm.), prepared from over two hundred dissections, illustrate the prevailing differences.