Rubus heterophyllus Willd., 1811

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 : 83-86

Rubus heterophyllus Willd.

In Die wilde Baumzucht: 413 ( Willdenow 1811).

Rubus plicatifolius Blanch. , Rhodora 8: 149 ( Blanchard 1906a).

— Lectotype, designated by Widrlechner (1998: 440): Wells Beach Depot , York County, Maine, 14.VIII.1905, W.H. Blanchard 477, BH.

NEOTYPE (here designated). — The photo in Bailey, Gentes Herbarum 1: 162 (1923). A drawing of the type was printed inBailey (1945: 773) .

FINDINGS

The original specimens that were attributed to this name at B have been lost. Bailey (1923) obtained a photo and a drawing of a specimen from B, which was only a young primocane. Bailey did not formally select it as a lectotype. Since it is no longer extant, we choose his photo as the new type. Though it was created from original material, the photo itself was made in the 20TH century and, thus, is a neotype (photo inBailey 1923: 162).

To begin with, we considered whether R. villosus Aiton and R. heterophyllus might be identical. Both have coarsely serrate leaves and bear weak prickles. However, the teeth of the leaves of R. heterophyllus are much sharper than those of R. villosus . Most notably, however, Willdenow wrote in his protologue that its leaves are glabrous, which does not correspond with the type of R. villosus . In addition, some of the terminal leaflets of R. heterophyllus are ovate and not shouldered like those of R. villosus . We conclude that they are not identical.

Many 19TH-century authors ( De Candolle & Sprengel 1820: 507; Steudel 1821: 706; Trattinnick 1823: 15; Dietrich 1837: 523; Heinhold 1840: 523) identified R. heterophyllus with R. villosus Torr. , while explicitly excluding R. villosus Aiton. By doing so, they supposed that Torrey had a different view of R. villosus than did Aiton. If they gave a specific reference, it was to Torrey (1819: 47), where he only presented a list of names, and did not claim that he differed from Aiton in his species concept. So their interpretations do not help clarify our understanding of R. heterophyllus .

Miquel (1867: 34) speculated whether R. heterophyllus might be R. thunbergii Siebold & Zucc. ( Siebold & Zuccarini 1835: 18) , heterotypic synonym of R. hirsutus Thunb. ( Thunberg 1813: 7) , which is certainly not correct: R. hirsutus has pinnate leaves, which are very hairy on both sides. Rydberg (1913: 473; 1915: 157) was mistakenly convinced that R. heterophyllus was R. villosus var. michiganensis F.W. Card ex L.H. Bailey (1898: 374) which is synonymous with R. roribaccus Rydb. ( Rydberg 1901: 498) ( Widrlechner 1998: 438) . That species has more substantial armature and softly pubescent leaves.

Bailey (1923: 162) supposed that R. heterophyllus might be R. recurvans Blanch. ( Blanchard 1904: 224) but proposed to drop the name as a nomen incertum ( Bailey 1945: 775) as he was inclined to do earlier ( Bailey 1923: 162). Palmer & Steyermark (1935: 568) were more positive about its identification as R. recurvans . But later, Bailey refuted this identification definitively ( Bailey 1945: 775), and correctly so. The type of R. recurvans is ( Widrlechner 1998: 456, lectotype): Middlebury, Addison County, Vermont, 21.VI.1899, E. Brainerd 24a and 24b, GH. If one only takes the pictures of R. heterophyllus into consideration, one might be inclined to accept the conspecificity of R. heterophyllus and R. recurvans . However, the protologue of the former makes such a conclusion impossible. Willdenow’s full protologue (1811: 413) (as translated by the senior author) reads as follows:

“12. Rubus heterophyllus , leaves 3- and 5-foliolate, glabrous, ovate-oblong, acuminate, deeply dentated, the stem aculeolate, petals spatulate-lanceolate.

Variable leaf bramble

The home land is unknown, probably North America.

A species two feet tall with many prostrate stems. The infertile stems are bluntly angled, with sparse fine hairs, with pointed, somewhat curved prickles provided. The fertile branches are round,glabrous below,with sparse short prickles, the upper part with fine hairs and almost without prickles. The leaves of the infertile stem are digitate,5-foliolate, those of the fertile stem 3-foliolate, the lateral leaflets deeply lobate.The leaves in the range of the flowers are simple or also 3-foliolate. The leaflets are ovate-oblong, long acuminate, deeply unequally dentate. The petioles of the infertile stem are prickly, those of the fertile stem with fine hairs.

The flowers are white, they come in July, the calyx and the pedicels are white hairy, the petals narrow spatulate, sometimes incised.”

The glabrous leaves and the phrases ‘two feet tall with many prostrate stems’ and ‘the petals narrow spatulate’ do not correspond with R. recurvans . Its leaves are hairy and, while it can sometimes produce arching primocanes with prostrate rooting tips, it never produces multiple, strongly prostrate stems. Its petals are obovate, but not conspicuously narrowed ( Widrlechner 1998; Widrlechner & Smith 2008). Willdenow’s description presents a plant which begins with young upright stems that soon become prostrate, so that the whole bush is not taller than two feet. This excludes the highbush brambles, and points towards a member of the Procumbentes, many of which take this form. So the identity of R. heterophyllus must be found there. After a thorough investigation of taxa within this section with the combination of stems that are low-arching at the base, leaflets with very sharp and jagged serrations, and narrow petals, only two close matches could be identified: R. plicatifolius and R. pronus L.H. Bailey (1943: 295 , holotype: BH, grassy weedy land by Mountain Lake, Garrett County, Western Maryland, Bailey 670).

Bailey distinguished between R. pronus and R. plicatifolius by the presence of stipitate glands in R. pronus , but Davis et al. (1968b) included them both among the eglandular Flagellares (synonym of Procumbentes), separating them by leaf shape. However, even the type of R. plicatifolius shows both types of leaves, leading us to compare the type of R. pronus with R. plicatifolius . After a close check of both taxa, we found that the leaves of the type of R. pronus are more pubescent abaxially, its pedicels are also pubescent and even more: they have some stipitate glands; so Bailey (1943) was correct in his observations. Other samples that are identified as R. pronus display the same characters. Thus, it seems that R. pronus is an unusual (probably rare) species with close affinities to R. plicatifolius . It cannot be established if the plant of R. heterophyllus had stipitate glands. However, Willdenow’s remark that it is glabrous refers better to R. plicatifolius (besides the fact that it is more probable that a rather common bramble was collected than a rare one).

Blanchard (1906a) based his publication of R. plicatifolius on plants from Wells Beach Depot. A specimen in BH was selected as the lectotype ( Widrlechner 1998: 440). Other specimens from this locality are conserved at NY. According to Blanchard (1906a: 150), the most striking characteristic of R. plicatifolius is “the plaited or ruffled margins of the leaves,” even adding that these make this species “unique among dewberries if not among blackberries in general.” It is precisely this characteristic which the picture of R. heterophyllus shows. It has the same curved tip of the leaflets as does R. plicatifolius , which is due to the plicate margins when dried, just as is the case with European species with plicate leaves, such as R. plicatus Weihe & Nees (1822 -1827: 14) and R. affinis Weihe & Nees (1822 -1827: 18).

There is another peculiar characteristic which merits attention. Willdenow (1811) mentioned that the calyx and the pedicels are white hairy. This has always been a stumbling block, for it suggests an affinity with a species of the Arguti, even if its leaves are glabrous. However, it supports the identification with R. plicatifolius , as Blanchard (1906a: 150) wrote in the protologue “sepals very pubescent or woolly”.

In sum, we find no traits that weigh against the conspecificity of R. heterophyllus and R. plicatifolius , but, more interestingly, there are such peculiar specifics that it appears impossible to not treated them as the same taxon. So the conclusion must be made that R. heterophyllus is the correct name of the taxon which has been known as R. plicatifolius until now.

Link (1822: 62) referred to the descriptions of Willdenow (1809: 38; 1811: 413) as R. tetraphyllus . It is not clear if this is a mere error or a conscious change. In the most extreme interpretation, R. tetraphyllus is but a later, homotypic synonym of R. heterophyllus . IPNI (2012) mentions it as ‘ Rubus tetraphyllus Willd. – Berlin. Baumz., ed. 2 413. 1811’, which is a clear mistake.