Flora of North America species comparison

Subspecies 2 (2 in the flora).

Chamaenerion angustifolium, which in North America is known mostly as fireweed, is a widespread circumpolar, circumboreal species that can be locally dominant, often in disturbed habitats, particularly following fires (T. Mosquin 1966, 1967; G. Henderson et al. 1979). In addition to producing numerous, highly vagile, comose seeds, aggressive rhizomatous growth enables it to survive clonally and spread rapidly after fires (L. E. Vodolazskij 1979). As succession proceeds, production of the familiar spikes of magenta flowers is inhibited, but fireweed often persists in non-flowering condition even in relatively mature forests, where it is able to sprout and spread quickly following disturbance. The species is one of the few within the genus and tribe to form true rhizomes (R. C. Keating et al. 1982), enabling it to colonize large areas very rapidly.

Polyploidy is well documented in Chamaenerion angustifolium, but unlike the similar situation in C. latifolium (T. Mosquin and E. Small 1971), diploid and tetraploid populations differ morphologically and have partially overlapping, but distinct, geographical ranges (Mosquin 1966; Chen C. J. et al. 1992). The diploid subsp. angustifolium occupies the northern part of its range, in North America across Canada and interior Alaska, and in Asia across Siberia and northern Europe, ranging southward at higher elevations. Farther south in Eurasia and North America, it is replaced by the tetraploid subsp. circumvagum (Mosquin 1966). Hexaploid (n = 54) populations have been detected only in Japan, and cannot be distinguished morphologically from tetraploids. B. C. Husband and D. W. Schemske (1998), Husband and H. A. Sabara (2004), and Sabara et al. (2013) have documented population segregation and mixing in contact areas between the two ploidy levels, including populations with a single cytotype, a high proportion of mixed populations, and presence of triploids (n = 27), but at low levels, indicating strong reproductive isolation between cytotypes.

The floral phenology of Chamaenerion angustifolium was described in 1793 by C. K. Sprengel (P. Knuth 1906–1909, vol. 2) as a classic example of protandry. The flowers are presented in tall spikes, opening from the base of the spike, initially with stamens erect and style sharply reflexed. By the second or third day after opening, the stamens reflex as the style straightens into the floral axis, the four lobes of the stigma spread open, and nectaries commence production. Bees start at the lowest flowers in search of nectar, moving up the spike until there is no more nectar, by which time they are well dusted with pollen, which they bring to the lower functionally (female) flowers of the next spike.

(Discussion copyrighted by Flora of North America; reprinted with permission.)

Genera 2, species 173 (2 genera, 43 species in the flora).

Epilobium and its close relatives have been recognized historically either as part of Onagreae (A. P. de Candolle 1828b), sometimes as a subtribe (É. Spach 1834–1848, vol. 4; J. Torrey and A. Gray 1838–1843, vol. 1), or as the distinct Epilobieae (R. Raimann 1893; P. A. Munz 1965; P. H. Raven 1976). The synapomorphies for Epilobieae as currently delimited include its highly condensed, heteropycnotic chromosomes with a base chromosome number of x = 18, sepals held erect or spreading (not reflexed) throughout anthesis, and the presence of a coma of hairs on the seeds (secondarily lost in some species). Molecular support for the tribe is strong (97–100% BOOTSTRAP support; D. A. Baum et al. 1994; R. A. Levin et al. 2004).

Endlicher established Epilobieae and included Oenothera within it; it is unclear how or whether his concept differed from Onagreae of A. P. de Candolle (1828b). É. Spach (1834–1848, vol. 4) recognized these genera as Onagreae and differentiated so-called sect. Oenotherinae from sect. Epilobieae, placing in the latter not only Epilobium and related groups, but also Clarkia and its segregates. J. Torrey and A. Gray (1840) excluded Clarkia from their Epilobiinae and also excluded Boisduvalia, a delimitation also followed by R. Raimann (1893) for his Epilobieae. Epilobieae did not assume its current delimitation, including only Epilobium and its close relatives, until the works of P. A. Munz (1941) and P. H. Raven (1964).

G. L. Stebbins (1971) and P. H. Raven (1976) considered the diverse chromosome numbers in Epilobieae and proposed that the species of Boisduvalia (now a section of Epilobium) with n = 9 or 10 represented the original base chromosome number for the tribe, and that these numbers were derived from x = 11 which is found in Circaeeae, Gongylocarpeae, and Lopezieae (W. L. Wagner et al. 2007). They proposed a series of aneuploid reductions from n = 9 or 10 to n = 6, followed by polyploidy to produce the array of numbers in Epilobium (n = 12, 13, 15, 16, 18, 30) and Boisduvalia (n = 9, 10, 15, 19). D. A. Baum et al. (1994) demonstrated that molecular data did not support that hypothesis and found that a monophyletic Chamaenerion (n = 18, 36, 54) forms a strongly distinct sister branch to Epilobium, within which sect. Epilobium (n = 18) is monophyletic and sister to the rest of Epilobium, including the former segregates Boisduvalia and Zauschneria. The data from Baum et al. suggested that Epilobieae are primitively polyploid, with numbers based on x = 18, which is unique in the family. Using comparable sampling and some additional genes, R. A. Levin et al. (2004) found strong support for the phylogeny proposed by Baum et al.

(Discussion copyrighted by Flora of North America; reprinted with permission.)