Panarctic Flora

8608 Antennaria Gaertn.

• Gaertn., Fruct. Sem. Pl. 2, 3: 410 (1791) nom. cons..
• Excluded: Antennaria dioiciformis, A. lanata.

Notes: Elven and Murray: Antennaria is a predominantly North American genus where Bayer (2006: 389-392) summarized the complicated evolutionary situation and its current taxonomic representation. The lowest chromosome number is 2n = 28 and Bayer considered the base chromosome number to be x = 14. There are several reports of numbers at intermediate ploidy levels (see, e.g., Chinnappa 1984), suggesting that the base number may be x = 7 and the 2n = 28 plants tetraploids.

Antennaria is dioecious. Nearly all the tetraploid plants are sexuals with both sexes present in the populations. Some sexuals are found also among the octoploids, but the majority of plants at higher ploidy levels (5x and upwards) are agamosperms and then usually only with females present in the populations. There is also a fairly widespread proposed species - A. isolepis (synonymized below with A. rosea subsp. pulvinata) - where only male (staminate) plants are known (see Hultén 1968a: 876), however that is done. Agamospermy is widespread in the northern species groups. Exceptions are A. dioica and major parts of the A. carpatica aggregate. Very complex allopolyploid origins have been suggested for some groups and species, e.g., A. alpina s. lat. and A. rosea s. lat. (see Bayer 2006: 391). This makes a more detailed subgeneric division inapplicable as the majority of the allopolyploids are suggested derived from crosses between species of the possible sections or series based on the sexual tetraploids.

Extensive description and naming of agamospecies has taken place, especially in North America and Greenland, by Ekman, Fernald, Greene, Malte (e.g., Malte 1934), M.P. and A.E. Porsild (e.g., Porsild 1950), Rydberg, and others. At the time of Hultén (1950), about 200 species had been described from North America. Hultén (1968a, 1968b) reduced this number appreciably for northwestern North America by lumping of morphologically similar agamic species. This region is also the only one where a synoptic survey of the taxa has been undertaken (Bayer 1993, with reference to most of his relevant earlier publications), exteded by Bayer (2006) to all of North America. Bayer (2006) accepted 45 species worldwide and 34 for North America including Greenland.

The relationships between the sexual and apomictic 'lines' are not always clear. The agamosperms may sometimes have arisen as taxonomic autopolyploids directly from a sexual plant (assumed for A. monocephala subsp. angustata from subsp. monocephala), more often as allopolyploids from hybridization between sexuals (in the A. alpina and A. rosea complexes). The variation in Antennaria has been approached in two ways in North America in recent times, by R.J. Bayer in numerous papers from the 1980s onwards (Bayer 1984, 1987, 1988a, 1988b, 1989a, 1989b, 1990, 1991, 1993, 1996, 2006; Bayer and Crawford 1986; Bayer and Stebbins 1981, 1987) and by J.G. Chmielewski (Chmielewski 1995, 1996, 1997, 1998; Chmielewski and Chinnappa 1988, 1990), most recently by Bayer (2006). The two approaches differ in some principles and in many details. One of these is that Bayer divides sexuals and agamosperms on different taxa (usually as subspecies), whereas Chmielewski relies more on morphology and is reluctant to divide along reproductive lines. The first approach would fit a situation where the agamosperms are taxonomic alloploids from two or more sexuals, but in many cases this is not documented. Both authors agree, however, in applying much broader species circumscriptions than done previously and to subdivide their wider, polymorphic species on subspecies. Our approach has been to follow Bayer and Chmielewski as closely as possible, to try to fit plants not yet treated (Eurasian and some Greenlandic plants) into their frameworks, and to try to 'smoothen' the discrepancies between the two approaches. Note, however, that these recent North American treatments of Antennaria by Bayer and Chmielewski (and also the North American treatment of, e.g., the Arnica angustifolia group) differ in principle from recent western European and partly Russian treatments of agamospermous groups. Most major European treatments of, e.g., Hieracium, Pilosella, the Ranunculus auricomus group, Rubus, Sorbus, and Taraxacum, have found the racial (subspecies) approach unsuitable for such groups and have rather focused on characterization and naming of each definable agamospeces ('microspecies'), e.g., Lundevall and Øllgaard (1999: Taraxacum) and Ericsson (2001: the Ranunculus auricomus group), an exception being Tyler (2001) applying a broader species circumscription in Pilosella (but more recently a narrow one in Hieracium s. str.). As the patterns of hybridization and agamospermy in Pilosella and Hieracium s. str. differ in principle, his different approaches are defensible. Even if we are skeptical to the application of subspecies, we favour the broader circumscriptions of Bayer and Chmielewski. However, much experimental work is needed before a similar approach can be applied to, e.g., Hieracium and Taraxacum, and there is also still some way to go in Antennaria.

The majority of arctic Antennarias fit into nine species or aggregates. The A. carpatica aggregate belongs to Bayer's group Leontipes, all the others to Bayer's group Catipes: A. dioica, A. microphylla, A. neglecta, A. rosea s. lat., A. media s. lat., A. friesiana s. lat., A. alpina s. lat., and A. monocephala s. lat. Urbanska has treated the comparatively simple A. carpatica aggregate in numerous papers (see references below). Bayer has treated the North American representatives of A. rosea s. lat. (Bayer 1989a, 1989b, 1990, 2006) and of A. friesiana s. lat. and A. monocephala s. lat. (Bayer 1993, 1996). Chmielewski has treated A. media s. lat. (Chmielewski 1997) and A. alpina s. lat. (Chmielewski 1998).

The A. alpina aggregate has been and still is especially problematic as it has a significant range also outside North America. Bayer (1993: 150): "however, a couple of problematic taxa, such as the A. alpina complex, are still in need of much more detailed investigation. Their taxonomy must still be considered provisional at this time". For this complex, several of the investigations of Chmielewski are relevant, especially Chmielewski (1998). The A. alpina aggregate can alternatively be considered as amphi-Atlantic or as a more widespread complex. Bayer (1993, 1996, 2006) and Chmielewski (1998) both preferred the latter course, whereas the former course has been the northwestern European tradition. The geographically more narrow concept leaves some northwestern North American plants unaccounted for, whereas the broader concept may make the complex inconsistent.

A second problem with both North American approaches is that - even if they are probably the most efficient approaches today - they cannot yet be applied circumpolarly. Collective treatment or as subspecies in North America but as species elsewhere would give a strongly biased impression of the evolutionary situation in different regions. Several parts of our Checklist treatment of Antennaria therefore remain provisional.