North American species of Rubus L. (Rosaceae) described from European botanical gardens (1789 - 1823) Author Van de Beek, Abraham Petenbos 8 3904 BN Veenendaal (The Netherlands) beekavd@xs4all.nl Author Widrlechner, Mark P. Iowa State University Departments of Horticulture and of Ecology, Evolution, & Organismal Biology 360 Bessey Hall, 2200 Osborn Drive, Ames, Iowa 50011 (United States) isumw@iastate.edu text Adansonia 2021 3 2021-04-12 43 8 1789 1823 journal article 7408 10.5252/adansonia2021v43a8 49f3bd37-fa27-45a4-90b2-1e80c7f027ac 1639-4798 4680762 Rubus inermis Willd. In Enumeratio Plantarum : 548 ( Willdenow 1809 ) . — Lectotype (here designated): B ( BW09891010 ) (selected by Monaterio-Huelin & Weber 1996 : 316, pro holotype). Rubus inermis Thunb., Dissertatio botanico-medica de Rubo : 7. — Lectotype (here designated): UPS ( UPS-THUNB 12270 ). FINDINGS Willdenow described R. inermis in his list of plants in the botanical garden of Berlin. He mentioned North America as its nativity. Though his description is rather extensive for that time, it is not sufficient per se for identification of a presently known species. Fortunately, there is a specimen of this taxon made by Willdenow in the Herbarium in Berlin (BW09891010). Monasterio-Huelin & Weber (1996: 316) selected it as the type. They considered it to be the holotype , but it is not certain that Willdenow solely relied on this specimen. Thus, it is better to consider it as a lectotype . Various authors ( Sudre 1908 -1913: 74; Focke 1914: 378 ; Bailey 1945: 846 ; Monasterio-Huelin & Weber 1996: 316) have identified R. inermis Willd. as a variety or variation of R. ulmifolius Schott. If this is correct, the plant cannot be of American origin, since R. ulmifolius is native exclusively to the Old World. So Willdenow or his original supplier may have made a mistake. However, in light of Willdenow’s explicit statement that the plant came from North America, we decided to revisit the possibility whether any Rubus taxon in North America (including Central America) could correspond to Willdenow’s plant. One motivation for doing so was that the acute 3-foliolate leaves found on the type specimen are atypical for R. ulmifolius and even more so is its tendency to bear simple leaves and compound leaves with normal central leaflets and two very small, lateral leaflets like ‘ears’ on the petiole. Willdenow did not give details of the inflorescence or flowers in his protologue. In his 1811 publication ( Willdenow 1811: 411 ), he wrote that the plant had not yet flowered. Willdenow’s successor in Berlin, Link (1822: 62) noted that the plant had perished but resembled R. caesius L. and associated it with R. flagellaris . Seringe (in De Candolle 1825: 559 ) conceived the plant as a variety of R. flagellaris . It is certainly not R. flagellaris , not even a variety of it, but it is clear that early interpreters sought its identity among small, trailing brambles. However, there is no similar plant in North or Central America, not even within other Rubus subgenera. Further investigations in R. ulmifolius Schott revealed that young or secondary sprouts sometimes produce leaves resembling those of Willdenow’s plant, with the same ‘ears’ and trilobate simple leaves. This is especially true of plants collected in the eastern part of its native range, leading us to the conclusion that R. inermis Willd. is a specimen representing an uncommon developmental phenotype of R. ulmifolius and by consequence taxonomically, but heterotypically, identical with R. inermis Pourret ( Pourret 1788: 326 ; lectotype (designated by Beek 1979 ): MAF[MAF 3168]; syntype : P[P02521232]). The same taxon was once again published with the same name by Thunberg four years later: R. inermis Thunb. ( Thunberg 1813: 6 , 9). Because Thunberg presented his R. inermis explicitly as a new species ( Thunberg 1813: 7 ), a type must be selected from his herbarium (UPS-THUNB 12270). This entity is taxonomically identical with R. inermis Pourret , but heterotypic. Rubus inermis Pourr. is an older legitimate name of R. ulmifolius . Because R. ulmifolius is a very common blackberry, and the name has been in use for a very long time, a proposal for its conservation is in preparation (cf. Beek 2016 ).