Rubus pensilvanicus Poir., 1804

Van de Beek, Abraham & Widrlechner, Mark P., 2021, North American species of Rubus L. (Rosaceae) described from European botanical gardens (1789 - 1823), Adansonia (3) 43 (8), pp. 1789-1823 : 77-79

Rubus pensilvanicus Poir.

In Encyclopédie méthodique, Botanique 6: 246 ( Poiret 1804).

R. abactus L.H. Bailey, Gentes Herbarum 2: 452 ( Bailey 1932).

— Lectotype (here designated), selected by James L. Reveal, 2013 (presumably based on a statement in Bailey [1945: 718]): Connecticut Hill , Ithaca, Tompkins County, New York, 1.VIII.1925, L.H. Bailey 2000 (lecto-, BH [ BH 000 078 925 , BH 000 078 928 ]) ( Fig. 6A, B ).

HOLOTYPE. — P00320321 (P-JU 14 335).

FINDINGS

The description by Poiret is based on a specimen in the collection of Jussieu. Consequently, since it is extant, this must be considered the holotype (P00320321). As was typical in that era, it only consists of a floricane.

This name was not used for any American bramble until Bailey received a photo of it from M.L. Fernald ( Bailey 1945: 704) and identified it with plants that he found at many locations in the eastern United States ( Bailey 1945: 702-704). Consequently, R. pensilvanicus suddenly became a rather common species. Bailey (1945: 699) treated R. pensilvanicus as a member of the Arguti. Fernald (1950: 861) and Davis et al. (1969b: 262) followed him in that regard. Notably, Alice et al. (2014) treated essentially all members of section Arguti under this name, believing it to be the oldest name for the pubescent, highbush blackberries lacking stipitate glands, which comprise this section.

However, Bailey based his identification only on a photo, one of insufficient resolution to notice the finest details. Thierry Deroin (P) checked Poiret’s type carefully and noted that it bears very short stipitate glands on its pedicels and also some on the inflorescence axis, along with a few longer stipitate glands. This observation was confirmed by high-resolution photos that he sent to us. Consequently, the type does not belong to the Arguti but instead to the Alleghenienses. In sharp contrast, Bailey (1945) made no mention of any glandular trichomes in his detailed description of R. pensilvanicus , consistent with his placement of it within the Arguti.

So, in a single publication, Poiret (1804) named two species that fall within the Alleghenienses: R. vulpinus , which clearly displays the stipitate glands that are a key character of that section and otherwise conforms to the plant widely known as R. allegheniensis , and R. pensilvanicus , with its sparse glandularity less typical of R. allegheniensis . Such a plant might even represent an intersectional hybrid or be an earlier name for R. abactus L.H. Bailey. Rubus abactus has been treated as a close ally of R. pensilvanicus within the Arguti (keying out in the same couplet in Davis et al. 1969b), but a careful examination of the type of R. abactus revealed that it also bears short stipitate glands on its pedicels, along with a few longer stipitate glands on its inflorescence axis, as well as having coarser, jagged leaf serrations.

If the type of R. pensilvanicus were more complete, we might be able to place R. abactus in synonymy under it, but there are at least two other options that cannot be fully evaluated without access to both well-developed primocane and floricane samples of R. pensilvanicus . First, it is possible that the R. pensilvanicus type represents a weak specimen of R. allegheniensis . Under suboptimal growing conditions, R. allegheniensis can produce smaller, weaker racemes that are atypical and may display few stipitate glands. Alternatively, in contrast to many apomictic North American blackberries, R. allegheniensis is typically sexual ( Aalders & Hall 1966; Thompson 1997) with the ability to hybridize with other sympatric taxa. If such hybrids involve Arguti, the progeny could present intermediate characteristics.

If one takes a very broad view of R. allegheniensis , with its considerable variation generated by being primarily a sexual diploid ( Aalders & Hall 1966; Thompson 1997), potentially either of Poiret’s names from the 1804 publication could be selected for designating the correct name. However, there are two strong arguments to select R. vulpinus . First, it is a typical representative of the species, while R. pensilvanicus is marginal by virtue of its ambiguous trait expression. Second, the name R. pensilvanicus has been in wide use in eastern North America since 1945 for non-glandular highbush blackberries, and its selection would cause considerable confusion if it would be applied to a common species in the same region, while R. vulpinus has never been so used. So R. vulpinus could be indicated as the correct name for the taxon which has long been called R. allegheniensis Porter. However , because the name R. allegheniensis has been in use for a long time for a very common species, one which is also widely cultivated, a change of name is not desirable. Thus, we plan to submit a proposal for its conservation.

Because of its glandular character, R. pensilvanicus does not belong to the Arguti. Consequently, if the Arguti are considered as one species, asAlice et al. (2014) did, the correct name of that species would be R. argutus Link , because this is the oldest of the legitimate names in that section.

A subsequent question is what must be the correct name of the taxon which was called R. pensilvanicus by Bailey and later authors who followed him. Neither Bailey (1945) nor Davis et al. (1969b) mentioned any synonym and we, too, could not find one. Consequently, a new name must be given to it. In order to avoid any further confusion, we describe this taxon as a new species of section Arguti and not as a new name for R. pensilvanicus sensu L.H. Bailey. We chose the name Rubus revealii sp. nov. to honor the late James Reveal ( Miller 2015), who contributed much to the knowledge of Rubus , not only of its American species but also its infrageneric taxa.

Figures

FIG. 6A. — Rubus abactus L.H. Bailey lectotype, primocane, L.H. Bailey 2000 (BH[BH 000 078 925]). Scale bar: 2 cm.