1. Echinacea angustifolia DC. (narrow-leaved coneflower)

Brauneria angustifolia (DC.) A. Heller

E. angustifolia var. strigosa McGregor

E. pallida var. angustifolia (DC.) Cronquist

Pl. 277 a, b; Map 1171

Plants with a usually elongated, vertical rootstock and often somewhat tuberous main roots, sometimes also with short, stout rhizomes. Stems 10–50(–70) cm long, mostly unbranched, sparsely to densely pubescent with stiff, spreading (loosely ascending elsewhere), pustular-based hairs. Leaves with the margins entire and usually pubescent with spreading hairs, the surfaces moderately to densely pubescent with stiff, mostly spreading (loosely ascending elsewhere), pustular-based hairs, strongly roughened to the touch, with 3(5) main veins. Basal leaves 5–25 cm long, the blade narrowly elliptic to lanceolate, mostly 5–20 times as long as wide, long-tapered or narrowly angled at the base. Stem leaves 3–10 cm long, linear to narrowly elliptic or lanceolate, otherwise similar to the basal leaves. Involucral bracts 6–12 mm long, the outer surface moderately pubescent with mostly pustular-based hairs, not glandular. Receptacle 2.0–3.5 cm in diameter, the chaffy bracts 9–14 mm long, hardened, usually dark purple toward the tip. Ray florets with the corolla 2.0–3.5(–4.0) cm long, 5–8 mm wide, mostly spreading at flowering, pink to purplish pink (rarely white elsewhere). Disc florets with the corolla 6–8 mm long, the tube yellow to green, the lobes usually dark purple. Pollen bright yellow when fresh (sometimes faded to pale yellow on herbarium specimens). Fruits 4–5 mm long. 2n=22 (2n=44 elsewhere). May–July.

Introduced, known only from the city of St. Louis (Montana to Wisconsin south to Texas). Railroads.

A number of historical specimens from Missouri attributed to this species were redetermined as E. pallida during Ronald McGregor’s (1968) research. These mostly represented fruiting specimens, immature plants in which the ray corollas were not fully elongated or drooping, or generally incomplete specimens. Steyermark (1963) cited a single specimen from a prairie in Shelby County to document the presence of E. angustifolia in Missouri, but this collection could not be located during the present study and likely was misdetermined. However, a collection made by Viktor Mühlenbach during his botanical survey of the St. Louis railyards does document the species in Missouri. In 1990, Michael Skinner of the Missouri Department of Conservation discovered plants on a dolomite glade in Ozark County that had somewhat shorter ray corollas than is usual for E. pallida, and for a time this population was thought to represent an extant native occurrence of E. angustifolia. However, recent critical examination of the sheet at the Missouri Botanical Garden Herbarium by Craig Freeman of the University of Kansas revealed that because of its drooping rays and pollen color, this population is better considered to represent atypical plants of E. pallida. The natural range of E. angustifolia approaches the Missouri border in adjacent Iowa, Nebraska, and Kansas, and it is in northwestern Missouri that this species should be sought in the future.

McGregor (1968) segregated var. strigosa for plants from Kansas, Oklahoma, and Texas with more appressed pubescence. In their morphometric analysis of the genus, Binns et al. (2002) were unable to discriminate these plants from those attributed to var. angustifolia in the portion of the species range where the two types co-occur. Cronquist (1955, 1980; Gleason and Cronquist, 1991) and Binns et al. (2002) also treated E. angustifolia and several other species as varieties of E. pallida, but this obscures the biosystematic differences between the taxa (McGregor, 1968; Baskin et al., 1993).