Flora of North America species comparison

The corollas of Gentiana andrewsii, G. austromontana, and G. clausa all remain completely and tightly closed but are pollinated by bumblebees, which force open the corollas. The fresh corollas of G. clausa are rounded at the summit, with the plicae concealed by the true lobes. In contrast, the fresh corollas of G. andrewsii and G. austromontana are more acute, with the plicae forming much or all of the visible summit. The corolla lobes of G. clausa are about as long and as wide as the free portions of the plicae, whereas those of G. andrewsii and G. austromontana are distinctly narrower than the plicae. Also, in contrast to those of other species of Gentiana in the flora area except for G. flavida and G. latidens, the calyx lobes of G. clausa when fresh spread almost horizontally rather than being nearly erect.

Reports of Gentiana clausa from Indiana to Missouri and elsewhere west of the range given here have been based on specimens of G. andrewsii var. dakotica, second- or later-generation plants derived from G. andrewsii × G. puberulenta, or other hybrids and introgressants. In these plants, in contrast to G. clausa, the sepals are lanceolate and nearly erect, and the lobes of the intact corolla do not entirely conceal the plicae.

Gentiana clausa is largely isolated ecologically and geographically, but a few hybrids with G. andrewsii, G. austromontana, G. decora, and G. saponaria are known.

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

The name Gentiana linearis var. lanceolata A. Gray was applied originally to plants referable to G. linearis, although the name G. rubricaulis was cited in synonymy. The name G. linearis var. latifolia was applied originally only to G. rubricaulis, but both of these names were applied subsequently to both that species and relatively wide-leaved specimens of G. linearis. This confusion has been responsible in some cases for the rejection of specific status for G. rubricaulis, and for erroneous reports of G. rubricaulis in New York and Vermont. Reports from Nebraska were based on an old misidentification of G. puberulenta. Reports from Saskatchewan were also based on misidentified specimens. Reports of G. rubricaulis in Maine and New Brunswick are correct, although these populations are disjunct by about 775 km from the easternmost populations in Ontario.

In marked contrast to all other species of Gentiana in eastern and central North America, including G. linearis, the involucral leaves of this species are strongly ascending and somewhat conduplicate as well as being wider, and envelop the proximal portion of the flower cluster.

In the vicinity of Lake Superior, where the ranges of Gentiana rubricaulis and G. linearis overlap, these species maintain their distinctness, with G. rubricaulis occurring in calcareous soils and G. linearis in granitic and similar strongly acid soils (J. S. Pringle 1968). A few hybrids of G. rubricaulis with G. andrewsii, which is likewise a calciphile, are known. These hybrids have been designated G. × grandilacustris J. S. Pringle.

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