4. Eupatorium hyssopifolium L. var. calcaratum Fernald & B.G. Schub. (hyssop-leaved thoroughwort)

Pl. 266 g, h; Map 1112

Stems 30–100 cm long, usually not hollow, moderately to densely short-hairy, dark purple to purplish brown, not glaucous, with small axillary branches or fascicles of axillary leaves at all or most nodes up to 1/2 as long as the main stem leaves. Leaves mostly in whorls of 4, those of the upper and lower nodes sometimes in whorls of 3 or opposite, rarely alternate, sessile or nearly so. Leaf blades 1–6 cm long, 0.5–6.0 mm wide, linear to narrowly oblong-elliptic or narrowly oblanceolate, more or less tapered at the base, tapered to a sharply pointed tip, the margins entire or less commonly with a few shallow teeth, often somewhat curled under, the surfaces glabrous or sparsely to moderately short-hairy along the midvein, also densely gland-dotted, with 1 main vein. Inflorescences terminal panicles, more or less flat-topped. Involucre 4–6 mm long, cup-shaped, the bracts lanceolate or narrowly oblong, mostly rounded or bluntly pointed at the tip, the margins thin and pale or transparent, faintly 1- or 3-nerved, densely short-hairy, green. Disc florets 5. Corollas 3.5–4.0 mm long, the surface often somewhat glandular, white. Fruits 2.5–3.5 mm long. 2n=20, 30, 40. August–November.

Uncommon in the eastern portion of the Ozark Division (eastern [mainly southeastern] U.S. west to Missouri and Texas). Savannas, upland prairies, and margins of sinkhole ponds; more commonly old fields, ditch banks, roadsides, and dry, open, disturbed areas.

This taxon was first reported from the open margin of a sinkhole pond in Howell County by Steyermark (1953), who considered it a relictual population disjunct from the Gulf Coastal Plain. More recently, however, it has become apparent that this species is naturally expanding its range northward into the Midwest, making use of disturbance corridors such as highways. It seems unlikely that it was present in Missouri prior to the twentieth century, and thus one might argue that it should not be considered native to the state. There is no evidence to suggest that it was introduced intentionally or unintentionally directly through human activities, and the northward migration of this species and others such as E. rotundifolium may have more to do with a warming trend in the region’s climate than with environmental perturbations.

Missouri plants described above are assignable to var. calcaratum, which is included within var. hyssopifolium by some authors (Gleason and Cronquist, 1991). It differs from the nominate variety in its shorter, narrower, mostly entire leaves, but there is a lot of intergradation between the two taxa. The two occupy similar ranges and habitats. More research is necessary to determine whether these morphological trends can be correlated with cytological or molecular characters.

A third variety that is accepted by most botanists is var. laciniatum A. Gray, which refers to more robust, apparently polyploid plants with the leaves mainly in whorls of 3 or opposite. Plants corresponding to this morphology are suspected to have arisen through past hybridization between E. hyssopifolium and one or more related species, but now reproduce themselves independently as a species with a unique range. They have been reported from as close as central Kentucky (Gleason and Cronquist, 1991). Their relationship to morphologically similar plants discussed in the next paragraph is unclear and requires further study. Sullivan (1978) reported hybrids between E. hyssopifolium and E. semiserratum from Alabama and Florida, some of which she suggested appeared morphologically similar to var. laciniatum. The possibility of these taxa hybridizing in southern Missouri should not be overlooked.

In Wayne County, on property owned by Dr. Robert Cacchione near Williamsville, a large, mixed population of E. hyssopifolium and E. altissimum in an old field has produced a number of morphologically intermediate hybrid plants. These putative hybrids, which appear to be fertile, will key to E. hyssopifolium in the key to species above, as they tend to produce leaves in whorls of 3 at many nodes. They are more robust plants than those of E. hyssopifolium, with broader leaves having relatively coarse serrations. These teeth tend to be deeper and more evenly distributed along the margins than in E. altissimum, but the leaves are not so large or broad as in that species. This hybrid situation, which is to be expected elsewhere in the southeastern Ozarks, has not been reported previously. It was brought to the attention of the Flora of Missouri Project by Bob Cacchione and Bill Summers, to whom thanks are extended. Voucher specimens are in the herbarium of the Missouri Botanical Garden.