5. Spiraea tomentosa L. (hardhack, steeple bush)

S. tomentosa var. rosea (Raf.) Fernald

Pl. 544 f, g; Map 2531

Plants shrubs, 0.5–1.2 m tall. Twigs reddish brown to orangish brown, usually rounded but often with fine longitudinal lines or ridges, finely and persistently woolly. Leaves sessile or short-petiolate. Leaf blades 3–7 cm long, ovate to lanceolate or oblong-elliptic, angled or tapered at the base, rounded or angled to a usually bluntly pointed tip, the margins finely and sharply toothed, the upper surface sparsely to moderately and finely hairy at maturity, green to dark green, the undersurface persistently and densely pubescent with tan to yellowish white, finely woolly hairs. Inflorescences terminal panicles of numerous flowers, longer than wide, narrowly ovoid to narrowly pyramid-shaped, often elongate. Hypanthia 1.0–1.5 mm wide, cup-shaped to somewhat conic, finely woolly. Sepals 0.5–1.0 mm long, triangular, bluntly to sharply pointed at the tip. Corollas not doubled, with 5 petals. Petals 1.5–2.0 mm long, pink (white elsewhere). Ovaries woolly-hairy. Fruits 2.5–4.0 mm long, woolly-hairy. 2n=24. June–August.

Uncommon, known thus far only from Dunklin and Madison Counties (eastern U.S. west to Minnesota, Arkansas, and possibly Louisiana; Canada). Mesic upland forests and margins of lakes.

Steyermark (1963) knew this species only from a historical collection from Crowley’s Ridge in Dunklin County. However, in 1998 an apparently native population was discovered in Madison County by David Lindsay, a naturalist at the S Bar F Boy Scout Ranch. Steyermark (1963) noted that although this species is widely cultivated elsewhere in the United States it is not commonly grown in Missouri as it requires moist, acidic soils.

Some authors have separated S. tomentosa into two varieties. Missouri plants correspond to the var. rosea, a widespread variant with slightly less densely flowered inflorescence branches. Plants from the northwestern United States correspond to var. rosea, with fewer flowers per cm of inflorescence branches. However, there is too much overlap between these variants in the eastern states to make their recognition feasible (K. R. Robertson, 1974).