Discussion

The North American and Central American species of Erigeron have been divided into sections (G. L. Nesom 1989c, 1990g, 1994b; Nesom and R. D. Noyes 1999), emphasizing variation in habit (especially taprooted versus rhizomatous and fibrous-rooted), vestiture, arrangement of heads in arrays and orientation before flowering (erect, nodding, or arching-pendent), behavior of ray corolla laminae (straight, reflexing, or coiling), cypsela and pappus morphology, and other morphologic features. The sequence and groupings of species treated here reflect significant modifications of earlier arrangements. G. L. Nesom (1989d) hypothesized that Trimorpha [Erigeron sect. Trimorpha (Cassini) de Candolle] is separate from Erigeron, more closely related to Conyza. Studies by W. Huber and colleagues (e.g., Huber 1993; Huber and Ö. Nilsson 1995) and R. D. Noyes (2000) have shown that Trimorpha species are closely related to those of sect. Erigeron and that both sections are relatively recently derived within the genus. As suggested by Nesom (1994b) and by Huber and Nilsson, and as discussed in detail and experimentally confirmed by Noyes, autogamous breeding systems apparently have arisen independently in groups of Astereae, including Trimorpha and Conyza, where the pistillate florets of a head are greatly increased in number (often outnumbering the bisexual florets), in multiple series, the inner sometimes with filiform, elaminate corollas, and the outer with reduced laminae. In the molecular analysis by R. D. Noyes (2000), Conyzinae comprises Erigeron, American Conyza, the four genera of the South American Leptostelma group, and the North American Aphanostephus; the cladistically basal and terminal taxa of the subtribe are members of Erigeron. Noyes (p. 107) observed that “strictly speaking, although the Conyzinae form a monophyletic group [with caveats regarding Old World Conyza], Erigeron is paraphyletic, as five other genera are derived from within it.” The molecular study included 46 of the 173 species treated here. Polyploidy is common among species of Erigeron, and agamospermy apparently is a common correlate of polyploidy, especially in odd-polyploid plants. Molecular phylogenetic data (R. D. Noyes 2000) indicate that agamospermy has arisen at least three times within the genus. In the descriptions and keys, some characteristics are assumed constant unless otherwise indicated (usually in parentheses); particular application of terms is discussed here. The indumentum of erigerons is often complex; in order to simplify descriptions, glabrous applies here only to absence of non-glandular hairs, eglandular to the absence of glandular hairs; a totally glabrous plant (in the usual sense) would be glabrous and eglandular. Petiole margins are eciliate or sparsely ciliate unless otherwise indicated. Leaf bases of most erigerons are broadened or not, not thickened and white-indurate. Margins of leaves in some erigerons are entire but for tiny callous enations that correspond to the callous tips of teeth on some leaves with serrate margins. Here, margins with such tiny enations are described as denticulate. Heads of some erigerons are “pseudodisciform” in the sense that the outer pistillate florets have relatively small, ± filiform laminae (such florets are technically “ray florets” even though their “rays” are inconspicuous) and the inner pistillate florets have no laminae on their corollas. The distinction between corollas without and those with laminae is sometimes arbitrary. Ray laminae are considered strap-shaped and spreading unless otherwise indicated. Descriptions of ray color as “blue” should be read as lavender-blue.

Discussion

The North American and Central American species of Erigeron have been divided into sections (G. L. Nesom 1989c, 1990g, 1994b; Nesom and R. D. Noyes 1999), emphasizing variation in habit (especially taprooted versus rhizomatous and fibrous-rooted), vestiture, arrangement of heads in arrays and orientation before flowering (erect, nodding, or arching-pendent), behavior of ray corolla laminae (straight, reflexing, or coiling), cypsela and pappus morphology, and other morphologic features. The sequence and groupings of species treated here reflect significant modifications of earlier arrangements. G. L. Nesom (1989d) hypothesized that Trimorpha [Erigeron sect. Trimorpha (Cassini) de Candolle] is separate from Erigeron, more closely related to Conyza. Studies by W. Huber and colleagues (e.g., Huber 1993; Huber and Ö. Nilsson 1995) and R. D. Noyes (2000) have shown that Trimorpha species are closely related to those of sect. Erigeron and that both sections are relatively recently derived within the genus. As suggested by Nesom (1994b) and by Huber and Nilsson, and as discussed in detail and experimentally confirmed by Noyes, autogamous breeding systems apparently have arisen independently in groups of Astereae, including Trimorpha and Conyza, where the pistillate florets of a head are greatly increased in number (often outnumbering the bisexual florets), in multiple series, the inner sometimes with filiform, elaminate corollas, and the outer with reduced laminae. In the molecular analysis by R. D. Noyes (2000), Conyzinae comprises Erigeron, American Conyza, the four genera of the South American Leptostelma group, and the North American Aphanostephus; the cladistically basal and terminal taxa of the subtribe are members of Erigeron. Noyes (p. 107) observed that “strictly speaking, although the Conyzinae form a monophyletic group [with caveats regarding Old World Conyza], Erigeron is paraphyletic, as five other genera are derived from within it.” The molecular study included 46 of the 173 species treated here. Polyploidy is common among species of Erigeron, and agamospermy apparently is a common correlate of polyploidy, especially in odd-polyploid plants. Molecular phylogenetic data (R. D. Noyes 2000) indicate that agamospermy has arisen at least three times within the genus. In the descriptions and keys, some characteristics are assumed constant unless otherwise indicated (usually in parentheses); particular application of terms is discussed here. The indumentum of erigerons is often complex; in order to simplify descriptions, glabrous applies here only to absence of non-glandular hairs, eglandular to the absence of glandular hairs; a totally glabrous plant (in the usual sense) would be glabrous and eglandular. Petiole margins are eciliate or sparsely ciliate unless otherwise indicated. Leaf bases of most erigerons are broadened or not, not thickened and white-indurate. Margins of leaves in some erigerons are entire but for tiny callous enations that correspond to the callous tips of teeth on some leaves with serrate margins. Here, margins with such tiny enations are described as denticulate. Heads of some erigerons are “pseudodisciform” in the sense that the outer pistillate florets have relatively small, ± filiform laminae (such florets are technically “ray florets” even though their “rays” are inconspicuous) and the inner pistillate florets have no laminae on their corollas. The distinction between corollas without and those with laminae is sometimes arbitrary. Ray laminae are considered strap-shaped and spreading unless otherwise indicated. Descriptions of ray color as “blue” should be read as lavender-blue.