Panarctic Flora

580213 Salix glauca L.

• L., Sp. Pl.: 1019 (1753). Lectotype: Sweden: "Lapland". Linnaeus, Fl. Lapp.: 363, t. 7, f. 5. 1737. (Jonsell and Jarvis 1994: 151).

2n= (1) 76 (4x). - Far East (N), Alaska, Canada. - Numerous reports, Canadian ones partly for S. seemannii.
(2) 95 96 (5x). - Far East (N), Alaska, Canada. - Suda and Argus (1969a); Petrovsky and Zhukova (1983a).
(3) >100. - Alaska. - A. Johnson in letter (1965) to G. Argus.
(4) 114 (6x). - Europe (N), Siberia (N), Far East (N), Alaska, Canada, Greenland. - Numerous reports, a Canadian one for S. cordifolia.
(5) 152 (8x). - Europe (N), Far East (N). - Several reports.
(6) 176. - Europe. - Wilkinson (1944, 1954).
(7) 190 (10x). - Far East (N). - Zhukova and Petrovsky (1977).
Suda and Argus (1969a) found 2n = 76, 95, and 114 in one population at Umiat in northern Alaska (within the range of subsp. stipulifera). Variation in ploidy, even at a local scale, is suggested.

Geography: Circumboreal-polar.

Notes: Argus and Elven: Salix glauca s. lat. is a very widespread and polymorphic species or species group. Several taxa have been accepted formally or informally, in northwestern Europe as two species (S. glauca and S. stipulifera, e.g., Rechinger 1964) or subspecies (e.g., Rechinger and Akeroyd 1993a; Elven 1994; Hämet-Ahti et al. 1998; Elven and Karlsson 2000; Elven et al. 2005), in Greenland as two subspecies (subsp. glauca and subsp. callicarpaea, e.g., Böcher et al. 1978), and in North America as several intergrading phases or varieties (Argus 1965, 1973 and later). Russian authors (e.g., Skvortsov 1968, 1999; Bolshakov 1992) have been reluctant to accept races. A reason for this divergence in opinions may be that the reported variation in this species is concentrated to North America (at least four races), Greenland (possibly two races, and northwestern Europe (two races), i.e., to non-Russian regions.

The majority of workers agree that the taxa are difficult to recognize because of their polymorphism and intergradation over large areas. It is worth notice that several ploidy levels are reported and that different (and different numbers of) diploid genomes may partake through recurrent polyploidy. Skvortsov commented that quite a number of formal and informal varieties have been proposed, all of very vague circumscription and of doubtful taxonomic value. Skvortsov and Yurtsev were reluctant to accept races, whereas Argus and Elven were more ready to accept them. Our decision to accept four races below is based on our ability to recognize four central tendencies, each with a distinct range, despite the wide area of overlap and intergradation between them. The reservation is taken that the variation in the Russian areas is largely unexplored.

Argus: Whether we should subdivide Salix glauca into races (subspecies or varieties) is a very difficult, but important, question. We could treat it simply as a highly variable species. But it would be difficult to justify including so much morphological variability within an undifferentiated species, especially since botanists working at a local level can see that their populations do not look like those from elsewhere. Yet there are virtually no invariable morphological characters to separate them (subspp. glauca, stipulifera, acutifolia, callicarpaea, and at least one more southern race). I do not have field experience with subsp. glauca but I know that the others, although distinctive in certain areas, intergrade with each other over vast regions. I cannot separate specimens of S. glauca from Siberia from those in Alaska. If this is true, I sometimes wonder how can we continue to maintain a classification that is based as much on geography as on morphology. The considerable variation in this species, as we all know, is the result of polyploidy (probably the repeated formation of polyploids combining different genomes), ecotypic variation, and environmental modification, all of which are confounded further by developmental variability and hybridization (especially with S. arctica). In my 1965 study I decided to treat the variants as informal phases, in the manner of some zoologists, and simply discussed the variability. This was not well received by botanists who want formal names but I think that it may be the best way to go. I agree with the argument used by Skvortsov in his Russian treatment. Also, I agree with Elven that if we do recognize infraspecific taxa, the rank of variety is not appropriate for wide-spanning geographical entities.

Some comments have indicated that perhaps subsp. callicarpaea was a recognizable taxon. It was also mentioned that there were problems of separating it from S. arctica in eastern Greenland, Iceland, and the Faroes, but in North America too there is extensive variation and some plants are putatively S. arctica x glauca. From Newfoundland, Fernald described a series of varieties in what is now treated as subsp. callicarpaea, based on leaf shape and pubescence, illustrating its variability in eastern Canada. On the western side of Hudson Bay and across northern mainland Nunavut and the Northwest Territories subsp. callicarpaea intergrades into subsp. acutifolia. If we recognize subsp. callicarpaea, then we must also recognize subsp. acutifolia, the extremes of both are equally distinctive. Then we are faced with the question of whether the low growing populations on the arctic coast of Alaska and in the Bering Sea area should be recognized, and we are back to the separation of subsp. glauca and subsp. stipulifera.

After all this weighing of pros and cons I believe that while it may be more accurate to treat S. glauca as a single, highly variable species, an argument could be made for recognizing four variable, intergrading geographical races. In the discussion for the Flora, we should attempt to say how the races differ but re-emphasize that even these characters are not definitive and that the races strongly intergrade. We should also note that many specimens (i.e., juveniles, plants in extreme sites, et al.) may not be identifiable to subspecies level. That would give botanists working in some regions the opportunity to use subspecific names for their races or to treat them all at the species level.

Elven: As for chromosome numbers, several are reported for all taxa but they need critical evaluation against vouchers (if available). The best supported number in northern Europe is the octoploid one (2n = 152), whereas there is support for both tetraploid and hexaploid numbers in North America (both subsp. acutifolia and subsp. callicarpaea) and for these and octoploid numbers in northeastern Asia. Decaploids are reported by Zhukova and Petrovsky (1977, as S. glauca s. lat.) from northeastern Asia and by Löve and Löve (1956b, as subsp. callicarpaea) from Iceland, but as all Icelandic material is currently assigned to S. arctica (Pálsson 2000) this last report is probably irrelevant for subsp. callicarpaea.

Argus (2010) retained the main features of the arguments above for his treatment in the Flora of North America but entered the races as varieties.

• 
  to be done
  Source: Hgrobe at Wikispecies