Rubus decumbens Thunb., 1813

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 : 86-87

Rubus decumbens Thunb.

In Dissertatio de Rubo: 5 ( Thunberg 1813).

Rubus arundelanus Blanch. , Rhodora 8: 176 ( Blanchard 1906b).

— Lectotype (here designated), selected by James L. Reveal, 2013: Casino in Kennebunkport , Me., 21.VII.1905, W.H. Blanchard s.n., BH ( BH 000 079 118 ) ( Fig. 9 ).

LECTOTYPE (here designated). — UPP ( UPP-THUN 12254 ), ‘ Rubus decumbens’, ‘e Niew Jersey. Hultgren.’ ( Fig. 8A, B ).

FINDINGS

Thunberg did not provide a full description of this species as a new taxon nor did he refer to any earlier publication. However, the short description in his 1813 overview is sufficient for valid publication, and according to his geographical overview, the species was native to North America ( Thunberg 1813: 9).

Two specimens of R. decumbens are present in Thunberg’s herbarium in Uppsala, numbers UPP-THUN 12254 and UPP-THUN 12255. On the reverse side of 12254 is written: ‘e Niew Jersey. Hultgren.’, and of 12255: ‘ Rubus foliis ternatis nudis, caule aculeato. Linn. flor. Suec. 410. Spec. 493. 4.’ The latter reference is to R. caesius L., but because it is only on the label and not in the publication, it is not of direct relevance for nomenclature. However, for the choice of a type, it might be a good argument to not select it. Other arguments are more decisive: UPP-THUN 12254 is explicitly from North America, which corresponds with the protologue. The same is true for the description of its leaves. Thunberg wrote that these are trifoliate and simple. Number 12254 has both kinds of leaves, but 12255 has only ternate ones. Taking all these factors together, we have chosen UPP-THUN 12254 as the lectotype.

The two specimens are not identical in other significant ways. UPP-THUN 12255 is rather glabrous; its inflorescences have only one or two flowers and none to a few weak prickles. In contrast, 12254 is hairier, both on the abaxial surfaces of its leaves and on its pedicels and sepals; its inflorescences bear five flowers, with rather strong prickles and stipitate glands. We speculate that 12255 might be a weak plant of R. flagellaris , but it is too poor for definitive identification.

The identity of our lectotype, UPP-THUN 12254 , is more important. From the way the flowering branches are arranged at the nodes of the floricane, it can be concluded that the sample was taken from a more-or-less horizontal stem. However, the cane bears declined prickles (some slightly decurved) that are up to 4 mm long, sometimes exceeding the diameter of the cane, rather than strongly decurved ones. Such a combination of prickle size and shape is more common in members of the Procumbentes with a mounding habit. Thunberg’s choice of the epithet, decumbens, a characteristic which he uses in the description as well, also points towards a plant with a low-arching to mounding habit.

Of the mounding Procumbentes found in eastern North America , the only taxon with this same combination of inflorescence form, leaf pubescence, and the presence of stipitate glands and prickles is R. arundelanus Blanch. ( Davis et al. 1968b) ; lectotype (here designated), selected by James L. Reveal, 2013: Casino in Kennebunkport, Me., 21.VII.1905, W.H. Blanchard s.n., BH (BH 000 079 118) ( Fig. 9 ).

The “next-closest” candidate, R. ithacanus L.H. Bailey , typically bears longer, more clearly racemose inflorescences and canes with few prickles ( Widrlechner & Smith 2008).

The type specimen of R. decumbens was collected by Matthias Hultgren, a Swedish Lutheran clergyman, who sent many plant specimens that he had collected in Pennsylvania, New Jersey, and Nova Scotia to European herbaria in the late 18TH century. He wrote on the label that the specimen was collected in New Jersey. We are currently trying to determine the extent of the native range of R. arundelanus and have identified representative collections from northern Massachusetts, Maine and points to the north and east in Canada. If our lectotype of R. decumbens was collected from New Jersey, it would represent a significant expansion of the known range of this taxon ( Davis et al. 1968b). We considered whether the label information for UPP-THUN 12254 might be incorrect; perhaps it was actually collected in Nova Scotia where Hultgren also gathered plants. However, Mats Hjertson (Uppsala) informed us that there is no past evidence that labels by Hultgren might be incorrect. It is more probable that the plant has become extinct in the type locality region. New Jersey and its neighboring states became well-settled in the 18TH and early 19TH centuries, a time when relatively few extant Rubus populations were sampled. We suspect that a better understanding of the native range will require extensive searching through herbaria, investigating specimens labelled as other members of the Procumbentes and perhaps even as R. allegheniensis , given its distinct glandularity. It is remarkable that Bailey also did not relocate R. arundelanus at its type locality in Maine in the 1930s ( Bailey 1945: 376). Perhaps it is a very vulnerable plant.

Figures

FIG. 9. — Rubus arundulanus Blanch. lectotype, W.H. Blanchard s.n., BH (BH 000 079 118). Scale bar: 2 cm.

FIG. 8A. — Rubus decumbens Thunb.lectotype (photo Mats Hjertson),UPP (UPP- THUN 12254). Scale bar: 2 cm.