Rubi Capenses: a further contribution to the knowledge of the genus Rubus (Rosaceae) in South Africa Author Beek, Abraham Van De text Phytotaxa 2021 2021-08-19 515 1 1 71 http://dx.doi.org/10.11646/phytotaxa.515.1.1 journal article 10.11646/phytotaxa.515.1.1 1179-3163 8061143 Series Pinnati Gust. ex A.Beek ser. nov . Type :— Rubus pinnatus Willdenow Diagnosis: —Primocane leaves 2- to 4-jugate pinnate, abaxially green, often only with some short hairs on veins, sometimes with longer shiny hairs. Comparison with other series: — Rubus ser. Pinnati differ from the other series with pinnate leaves by its abaxially green leaves. From R. ser. Rigidi Gust. ex A.Beek (newly described below) also by its shorter petals ( 3–6 mm vs 6–12 mm ) and from R. ser. Apetali A.Beek (newly described below) by its longer petals ( 3–6 mm vs 0–3 mm ). Notes: — Gustafsson (1934) described Rubus subsect. Pinnati Veri Gustafsson (1934:9 , 55). Because of the addition ‘Veri’, the publication is invalid. Rubus ser. Pinnati are validly described here as a series, a level that corresponds better with a closed group of taxa. Specimens that have previously been identified as R. pinnatus are not homogenous. Differences in pilosity, prickles, colour of the ripe fruits, and other characteristics cause so much diversity that it is rational to break the complex up into several entities. Five of these will be presented in this paper, but probably new taxa will be added after further investigations in the rest of Africa. Plants from the Western Cape have dense short hairs on the primocane. Specimens from eastern regions have (almost) glabrous stems, as well as other differences. Other glabrous and pruinose specimens from the Eastern Cape , often with strong compressed prickles and sometimes a few stipitate glands, might belong to other taxa, but the clarification of these must wait for an assessment of the eastern brambles. A salient taxon characterised by many long stipitate glands was found near Hogsback in the Eastern Cape . Within the Western Cape group there is a further level of heterogeneity. Two main forms can be distinguished. The common form has adaxially glabrous leaves and abaxially only some very short hairs on the main nerves, at first sight being glabrous; further: prickles of the flowering branch moderately strong and numerous; petiolules of the lowest leaflets 4–10 mm , those of the central leaflet 21–24 mm ; serrature rather obtuse or fine and slightly periodical; inflorescence pyramidal, often hidden in the leaves; pedicels and sepals greyish green. The other form has adaxially some strigose hairs on the leaves and abaxially long white hairs on the veins, at first sight hairy; further: prickles of the flowering branch stronger and more numerous; petiolules of the lowest leaflets 0–4 mm , those of the central leaflet 6–17 mm ; serrature very sharp and short periodical; inflorescence narrow and leafless; pedicels densely grey tomentose and hairy. The first form corresponds with the descriptions of R. pinnatus by Willdenow (1799) and by Chamisso & Schlechtendal (1827) , Ecklon & Zeyher (1836) , and all later authors. Ecklon & Zeyher (1836) gave a new name to the second form: R. pappei Ecklon & Zeyher (1836: 263) . Only a few specimens of it were collected in South Africa . Later authors lumped both as R. pinnatus . The forms are so different that the decision of Ecklon & Zeyher (1836) , who collected and distributed both, seems to be correct. The decision seems to be simple. However, the specimen of R. pinnatus in B-W belongs to the second form ( Fig. 1A ). If this would be the type , R. pappei and R. pinnatus would be identical and the common plant, which is generally called R. pinnatus , should have another name. This unfortunate situation is a reason to consider the typification of R. pinnatus carefully. Choosing another type than the specimen in B-W could be the best solution. The specimen in B-W is in serious conflict with the protologue. The leaves are not glabrous on both sides. For the interpretation of ‘glabrous’ the text of the protologue can be compared with the description of Willdenow (1799) for R. rosifolius . He writes there that its leaves are ‘subpilosa’ below. Of course, ‘subpilosa’ indicates more hairs than the simple ‘glabrous’. The specimen of R. pinnatus in B-W is much more hairy than R. rosifoliu s, so that Willdenow must have used another specimen when describing R. pinnatus . There is another specimen from the Willdenow collection in the herbarium of H.G. Bongard in Saint Petersburg in LECB (LECB0001886). However, this does not help very much for it is a sample of R. laciniatus Willdenow (1806 : t. 82), which is even more conflicting with the protologue. Consequently another specimen must be designated as type, which must be a neotype , because of lack of original material that corresponds with the protologue. The specimen of Mundt & Maire in B is a good option for this (‘ In promontorio Bonae Spei prope Hangklipp in nemoribus’, August 1821 , Mundt & Maire 225 , B101068581 : https:// herbarium.bgbm.org/object/ B101068581 ). Chamisso & Schlechtendal (1827) compared this with a specimen of Willdenow and saw no differences, though it is the common, glabrous form. They probably had another plant of Willdenow on their desk than the specimen that is now in B-W. Designating this specimen as the type of R. pinnatus supports continuity in the use of the name. Willdenow (1799) had only a dried sample (he writes ‘v.s.’ = vidi siccatam ). He does not give an indication of the locality of R. pinnatus ; after ‘habitat’ nothing is filled in. Most botanists in the beginning of the nineteenth century (e.g. Poiret 1804 , Miller 1807 ) only repeat Willdenow (1799) and do not give a country of origin. The first reference to a locality of R. pinnatus is found in a list of plants of the hortus in Paris by Desfontaines (1804) . He gives ‘Île de France’ (= Mauritius ). His colleague in Paris, Poiret (1804) , does not follow him. He does not mention a R. pinnatus from the garden in Paris and also not from the ‘Île de France’. He has another bramble from that island (see below under R. apetalus about his mistake), R. apetalus , which has also pinnate leaves. Probably Desfontaines (1804) used the same epithet as Willdenow (1799) , but meant another taxon. He does not refer to Willdenow (1799) as he does with other species in his book. FIGURE 1. Rubus pinnatus . A. Specimen in B-W, leaf detail abaxially. B. Primocane with leaves. C. Primocane. D. Primocane. E. Leaves. F. Leaves abaxially. G. Inflorescence. H. Inflorescence; I. Pedicels. Photographs: A. Katharina Rabe ; All others by A. van de Beek. Another indication of the nativity of R. pinnatus is given by Aiton (1811) . He writes that R. pinnatus was introduced into Kew Gardens in 1789, and that its origin is from St. Helena and Cape Good Hope, and gives it the English name ‘Wing-leaved Cape Bramble’, so indicating a relation of R. pinnatus and the Cape. Chamisso & Schlechtendal (1827: 19) were convinced a plant, which they got from Mundt und Maire from South Africa , is identical with Willdenow’s R. pinnatus , and, based on this conviction, they concluded that South Africa must be the homeland of R. pinnatus . Because Aiton (1811) also mentions, next to the Cape, St. Helena as native country of R. pinnatus , this island cannot be excluded from being the origin of R . pinnatus . According to O’Meara (1825) it was introduced in St. Helena , where it was very invasive ( Antommarchi 1825 ). All samples from St. Helena that I have seen are identical with the specimen of R. pinnatus in B-W. Aiton (1811) writes that Banks introduced it to Kew in 1789. Donn (1811: 144) enlists the species also for the botanical garden in Cambridge, where it was introduced in 1793. Aiton (1811) does not cite Willdenow (1806) as he usually does in his book, and the description is similar, but not identical. It follows that he connected the species not to Willdenow, but considered it as related to Kew Gardens. Samples from St. Helena in L, LY, and P are identical with the specimen in B-W. Many specimens of this island were collected by Cuming, who lived in the Philippines for a long time. This was the reason that some authors believed that R. pinnatus was found there. This was supported by labels of Cuming 2455 , whereon next to the name of the species only the name of Cuming and his residence ( Philippines , Manilla ) are written (P00682403; P00682406). Another label in P indicates the locality where it was found (P04203582), and a label with full data is at L.1918962. Focke already corrected this confusion on one of the sheets (L.1918962, the collection number is Cuming 2455 , not 2433 ). The specimen of Cuming 2455 in LY was described independently as R. cumingii Gandoger (1918: 24 , see below). Another specimen form St. Helena is P02579475. According to a note on the sheet of Willdenow’s plant in B-W, he received the specimen from Kerner. Kerner was the conservator of the botanical garden in Stuttgart, as Willdenow was in Berlin . It is not indicated where Kerner got his plant from, and it is not even certain that the indication is correct. The conclusion must be that there is a common species in the Western Cape, which corresponds with the diagnosis of R. pinnatus by Willdenow (1799) , and a more hairy taxon, which is common on St. Helena and hardly found in the Cape to which the present specimen in B-W belongs. It is not clear where this specimen and the specimen that Willdenow (1799) used for his description were collected.