Abraham Van de Beek, Mark P. Widrlechner
Adansonia 43 (8), 67-98, (12 April 2021) https://doi.org/10.5252/adansonia2021v43a8
KEYWORDS: Rosaceae, Rubus, Europe, batology, European gardens, lectotypifications, neotypifications, new species
Native Rubus were brought from North America to European botanical gardens in the 18th and early 19th centuries, and some were described as new species. Limited knowledge of Rubus in that era, typically combined with brief descriptions, led to many incorrect interpretations, causing much confusion (especially in North America). We established a project to clarify the identity of these confusing species by reviewing types and descriptions along with comparative work involving modern collections. We identified seven putative North American Rubus species named from early European garden collections: R. villosus Aiton, R. vulpinus Poir., R. pensilvanicus Poir., R. flagellaris Willd., R. inermis Willd., R. heterophyllus Willd., and R. argutus Link, and four other names in the early European botanical literature of possible North American origin: R. decumbens Thunb., R. inermis Thunb., R. enslenii Tratt., and R. floridus Tratt. We affirmed the current applications of R. flagellaris, R. argutus, and R. enslenii. Rubus villosus is identical with R. leviculus L.H. Bailey, while the taxon to which R. villosus was generally applied in the 19th century, R. allegheniensis Porter, appears to be identical with R. vulpinus. Rubus heterophyllus Willd. was identified as an earlier name for R. plicatifolius Blanch. Rubus inermis Willd. is identical with R. ulmifolius Schott and must have originated in the Old World. Thunberg's other American Rubus, R. decumbens, can be identified as R. arundelanus Blanch. Rubus floridus remains a puzzle. If its type represents a sample of a determinate flowering branch from a large, mounding dewberry, it is closest to R. grimesii L.H. Bailey, but it could also have been taken from a side-branch from an unknown upright blackberry. Another name for which identity could not be definitively established is R. pensilvanicus. It is allied with R. allegheniensis, but its type is insufficient to determine if it is an extreme form of that species, a related taxon, or a hybrid of it with a species of section Arguti (Rydb.) L.H. Bailey. We propose the name, R. revealii sp. nov. for the corymbose to weakly racemose Arguti previously considered as R. pensilvanicus sensu L.H. Bailey. Herein, we lectotypify R. abactus L.H. Bailey, R. arundelanus, R. decumbens, R. grimesii, R. inermis Thunb., R. invisus (L.H. Bailey) Britton, and R. vulpinus, and designate neotypes for R. argutus and R. heterophyllus.