Panarctic Flora

630638 Oxytropis uniflora Jurtz.

2n= 16 (2x). - Far East (Wrangel Island). - Zhukova in Yurtsev (1988, from type).

Geography: Asian Beringian: RFE.

Notes: Murray and Elven: Yurtsev (PAF proposal, comments) accepted eight species of sect. Gloeocephala to occur in the Arctic: four in northern and northeastern Asia, one amphi-Beringian, and three in northwestern North America. American authors have in the last 25 years mostly included the North American plants in one species, by Scoggan (1978c) as varieties of Oxytropis leucantha (almost certainly erroneous because O. leucantha is not 'glutinous' and also differs in numerous other assumed important features), by Welsh (1991, 2001) and Gillett et al. (2007) as varieties of O. borealis, and by Cody (1996) as synonyms of O. viscida. The exception is Porsild and Cody (1980, probably mainly Porsild, followed by Yurtsev) who recognized four species: O. glutinosa, O. hudsonica, O. sheldonensis (non-arctic), and O. viscida. There is no current North American treatment supporting Yurtsev's proposal but neither is there any supported alternative. Our own opinion concerning the northwestern North American parts is that a collective O. borealis or O. viscida is too inclusive.

Yurtsev's comments to this aggregate are here given in full except for exclusion of some polemics and a few orthographic corrections. This is done because they are the only justification for his treatment yet available in English. Some obscure words we have not been able to interpret are left in.

Yurtsev: The natural, whole ramification of sect. Orobia Yurtsev probably meant [Gloeocephala] is a development from an Orobia stock, marked with appearance - up to dominance - of viscid (verrucose) glands; their distribution on different organs, size and abundance, provide important characters for the taxonomy and phylogeny of the section. One of the not very many cases of intense allopatric speciation and race formation in the Arctic. The choice of O. borealis as the basis of possible subspecies combinations is formally correct but it ignores the relationships within the section. The section includes at least three natural groups: (1) The nearly circumpolar ?, the section is absent from nearly half the Arctic, from Hudson Bay east to Taimyr, 80W to 80E O. viscida aggregate whose members are most similar to those of sect. Orobia (numerous but not large flowers, red/blue or sordid white corolla with narrow banner and wings, narrow asymmetrical calyx, dorsal vein of legumes not profound), very minute glands. 2n = 16 (2x) or 32 (4x). Boreal-hypoarctic morphotype. (2) Eastern Siberian O. middendorffii s. lat. with nine major arctic or metaarctic races (subspecies). An arctic morphotype. The core of this alliance is formed by three to four races: subspp. middendorffii, orulganica, submiddendorffii. The eastern populations of subsp. orulganica were recognized as subsp. albida Jurtz. Flowers are much larger, with very wide banner and wings, teeth of calyx almost filiform, equilong with or even longer than tube. Calyx tube along with legumes are densely black (or black and white) pilose, with glands only on teeth and at the base of style. Dorsal vein is profound, often with very narrow dissepimentum. Leaflets are glandulose only underneath. 2n = 48 (6x). Stipules largely ovate, ciliate, sticky-yellow due to numerous glands. Glands large, yellowish. In subsp. submiddendorffii racemes, however, black and white spreading villose, and stipules not yellow, calyx teeth a bit thicker and shorter. Subspecies coerulescens, perhaps, most approaches a presumed ancestral type of O. middendorffii s. lat.: glands are minute, greenish and seen only on some parts of a plant (e.g., teeth of calyx, base of styles); on bracts, stipules, and bases of leaflets are clavate colourless processes characteristic of many species of sect. Orobia, but in sect. Gloeocephala these are met with only in this subspecies and its closest relative subsp. jarovii. These two taxa might be considered one species with two subspecies. At last a separate position in O. middendorffii s. lat. is occupied by subspp. schmidtii, anadyrensis, and trautvetteri with strongly glandulose calyx and legumes; leaflets glandulose on both surfaces. All the above taxa differ in several characters and should not be ignored. Oxytropis middendorffii s. lat. is an eastern Siberian arctic (and subarctic-alpine) group, derivative of the O. viscida aggregate. (3) Oxytropis borealis and, in all probability, O. hudsonica are the North American derivatives of the O. viscida aggregate somewhat similar to O. middendorffii s. lat., e.g., in large campanulate calyx, densely pilose calyx tube and legumes lacking glands, and flat wider leaflets, but largely differs from them in compact racemes, spreading sordid white and black pubescence of calyx and legumes, shorter teeth of calyx (less than half of tube). In O. borealis 2n = 48 (6x) as in O. middendorffii s. lat. As to O. glutinosa with lax racemes and long teeth of calyx and 2n = 16 (2x), it hardly belongs to a O. borealis aggregate. So it is hardly rational and correct to include the members of the O. viscida aggregate or O. middendorffii s. lat. in an O. borealis aggregate. It is interesting that in the easternmost Chukchi Peninsula O. middendorffii s. lat. is replaced by the mainly North American O. borealis.

Murray and Elven: We would strongly prefer some experimental support for this account but have no alternative that would do justice to the Russian variation. To lump the North American variation (following Welsh 2001) while accepting a split of the Russian one is unacceptable. Yurtsev's proposal is therefore largely accepted, waiting for a biosystematic investigation.

Welsh (2001) accepted three "viscid-glutinous" taxa from Alaska and the Yukon Territory as varieties of O. borealis (in addition to a stray record of var. hudsonica): var. borealis (incl. O. glutinosa as typified), var. viscida (incl. O. viscidula, O. glutinosa p.p.), and var. sulphurea (incl. O. verruculosa). His key is not very effective.

We do not include Yurtsev's proposed aggregates as their circumscription is not clear. He probably meant his O. viscida aggregate to include species 37, 38, 39, and 40, his O. borealis aggregate to include species 41 and 43, his O. middendorffii aggregate to include species 44, whereas O. glutinosa was left unassigned.