Flora of North America species comparison

Rubus ulmifolius can be distinguished, especially from the closely related R. bifrons and R. vestitus, by its strongly pruinose stems, finely serrate leaflets, and lack of glands throughout. Unlike R. ulmifolius, strongly pruinose native Rubus species lack relatively large and showy pink petals. Some new stems developing from tip-rooting, and early leaves on such stems (especially in shade), are not whitened abaxially and are tomentose. Such unusual stems develop typical leaves and surfaces in parts formed later.

There has been nomenclatural confusion over unarmed plants of Rubus ulmifolius (see E. Monasterio-Huelin and H. E. Weber 1996). Rubus inermis Willdenow is an illegitimate later homonym of R. inermis Pourret, and the type specimens for both names are of poor-quality. If an unarmed variety of R. ulmifolius is to be recognized, such as occurs in California, the name var. anoplothyrsus Sudre should be used instead of var. inermis (Willdenow) Focke.

Rubus ulmifolius is attractive and is potentially weedy but relatively uncommon, found primarily in coastal California from the San Francisco Bay region southward. It probably no longer persists in New Jersey, last collected there in 1897, and it is rare in Nevada and Oregon. Rubus ulmifolius likely hybridizes with R. bifrons where they co-occur; it is one of relatively few diploid and sexually reproducing species of European blackberries.

(Discussion copyrighted by Flora of North America; reprinted with permission.)

Species 250–700 (37 in the flora).

Molecular data (L. A. Alice and C. S. Campbell 1999) have shown Rubus to be monophyletic when including Dalibarda (R. repens). These data also show that long-recognized infrageneric taxa, such as the raspberries (subg. Idaeobatus Focke) and dwarf raspberries [subg. Cylactis (Rafinesque) Focke], may not be monophyletic. Although species are not organized here according to subgenera, they would belong to L. H. Bailey’s (1941–1945) subgeneric treatment as follows: flowering raspberries or thimbleberries [subg. Anoplobatus (Focke) Focke, species 3, 9, 18, 22, 23], snow raspberries [subg. Chamaebatus (Focke) Focke, species 19], cloudberries or baked-apple berries [subg. Chamaemorus (Hill) Focke, species 7], dwarf raspberries or plumboys (subg. Cylactis, species 2, 16, 26, 29, 31), raspberries (subg. Idaeobatus, species 11, 13, 14, 17, 20, 21, 24, 28, 33), and blackberries (subg. Rubus, species 1, 4–6, 8, 10, 12, 15, 25, 27, 32, 34, 35, 36, 37). Bailey did not include Dalibarda repens (30. R. repens) in Rubus, but W. O. Focke (1910) included it in subg. Dalibarda (under the illegitimate name R. dalibarda Linnaeus).

Rubus, especially the blackberries, presents some of the most difficult species-level problems, because of polyploidy, apomixis, and hybridization. As a result, differences of opinion on the number of species to be recognized from a given region can vary tremendously (for example, treatments in M. L. Fernald 1950 and H. A. Gleason and A. Cronquist 1991 for the northern half of the eastern United States plus parts of southeastern Canada, the former recognizing 205 species, the latter 25). A relatively conservative approach is taken here, one with much precedent (for example, A. E. Radford et al. 1968; R. G. Brown and M. L. Brown 1972; R. K. Godfrey and J. W. Wooten 1981; E. G. Voss 1972–1996, vol. 2; Great Plains Flora Association 1986; Godfrey 1988; Gleason and Cronquist; A. F. Rhoads and W. M. Klein 1993; E. B. Smith 1994; R. S. Mitchell and G. C. Tucker 1997; R. P. Wunderlin 1998; D. W. Magee and H. E. Ahles 1999; T. S. Cooperrider et al. 2001; R. L. Jones 2005; A. S. Weakley 2012; Weakley et al. 2012).

The decision to treat Rubus conservatively herein is due in part to the broad geographic distribution of many species in which quantitative characters are highly variable and the general absence of empirical data demonstrating that minor morphological variants are genetically based and not environmentally plastic. There is an overall lack of comprehensive herbarium specimens (for example, primocane, floricane with flowers and/or fruits, and habit data) making study difficult. Regarding Rubus, R. K. Godfrey (1988) wrote, "oversimplification appears to be the only way to achieve a practicable solution to the dilemma."

Other genera present a similar predicament by containing a large number of microspecies, such as Alchemilla, Crepis, Hieracium, Oenothera, and Taraxacum. W. Dietrich et al. (1997) stated for Oenothera taxonomy, and equally applicable to Rubus, that "immense splitting would result in a taxonomic system where even the specialist would lose the overview." See P. A. Rydberg (1913c), L. H. Bailey (1941–1945), P. D. Strausbaugh and E. L. Core (1978), H. A. Davis (1990), and M. P. Widrlechner (1998) for Rubus treatments that recognize many species.

Definitions for two morphologic terms primarily associated with Rubus are: primocane, a biennial or perennial stem prior to being developed enough for flowering, with flowering not occurring until at least its second season of growth, and floricane, a biennial or perennial stem after it has begun flowering, which does not occur until at least its second season of growth (R. W. Kiger and D. M. Porter 2001). Parcifrond is an elongate, sparsely leafy, sterile branch arising from a floricane proximal to its flowering branches. Novirame is a fertile (flowering, fruiting) branch on a primocane. Descriptions in this treatment are based on primocanes and floricanes, and exclude parcifronds and novirames.

In key leads and descriptions in this treatment, references to leaves are for those on primocanes, unless otherwise indicated. Also, for stems unsupported by other vegetation, measurements of vertical distances above the ground are given, not stem lengths. For example, stems of Rubus flagellaris may be as much as 3 dm above the ground even though the stems may be several meters long. Rubus inflorescences have been described as cymose or racemose. The inflorescences are determinate (thus cymose), with the terminal flower opening first. Flowering then proceeds acropetally (J. C. Guillard 1857; H. W. Rickett 1944), a pattern that is more typical of racemes. In the treatment here, inflorescences are described as cymiform, umbelliform, paniculiform, racemiform, or thyrsiform, these characterizations based on what the inflorescence most closely resembles. Also, inflorescence descriptions here refer to the flowering structures that generally terminate a branch distal to the leaves on that branch, including undivided, smaller, bractlike leaves. In plants with racemiform, paniculiform, or thyrsiform inflorescences, smaller axillary inflorescences containing one to few flowers are common and typically occur just proximal to the main inflorescences. Although most Rubus species in the flora area bear inflorescences that terminate a branch, such inflorescences often can appear to be axillary if terminating very short, leafy, axillary branches. Especially in subg. Rubus, terminal inflorescences almost always form on lateral branches, but occasionally form on a distal extension of the primocane of the previous year. In the flora area, native species in subg. Rubus have shorter floricanes, generally less than 30 cm; most non-native members in the flora area can have floricane branches exceeding 1 m.

Rubus shows diversity in armature and glands. Prickles, which protect the plant and aid some species in climbing, range from stout and broad-based to weak and narrow-based. Bristles may be stiff or weak. Prickles and bristles range from erect to retrorse. Generally, plants with sparse armature have spaces between the prickle bases or bristles that are more than 2–3 times the length of the prickles or bristles; plants with dense armature have spaces that are equal to or less than lengths. Glands vary significantly in size and shape and can be stalked or sessile. Glandularity is not always obvious; glands are sometimes hidden by other hairs or can be very small. Sessile glands can be less than 0.2 mm long or wide, requiring a good lens to be viewed; they are spheric to obovoid, whereas stipitate glands can be spheric to obovoid or discoid to cupulate.

Cultivars of blackberries and raspberries are grown for their edible fruit, and other Rubus taxa are grown for their ornamental value. Cultivars of blackberries have parentages or origins that include species native to North America and Europe, as well as red raspberry (for example, R. idaeus subsp. idaeus and subsp. strigosus) (J. R. Clark et al. 2007). Cultivars of red and black raspberries are primarily derived from R. idaeus and R. occidentalis, respectively. Purple raspberry cultivars are derived from the crossing of red and black raspberries. In both black and red raspberries, a lack of anthocyanin pigments can lead to the production of yellowish fruit. Cultivars of R. idaeus with yellowish fruit are sold as golden or amber raspberries. The Logan and Boysen cultivars are widely grown for their excellent fruits. Both cultivars are polyploid (6x and 8x respectively) involving R. ursinus as the maternal parent. In Loganberry, the paternal parent is R. idaeus subsp. idaeus, while Boysenberry has a more complicated ancestry.

Due to horticultural or agricultural interest, it can be expected that more Rubus species will be introduced into the flora, with some naturalizing. Rubus tomentosus Borkhausen var. canescens (de Candolle) Wirtgen, a European species similar to R. bifrons and R. vestitus but much more pubescent than either, was reported growing wild in a small area in Randolph County, West Virginia (E. E. Hutton and R. B. Clarkson 1961), although more recently it has not been confirmed as naturalized there (P. J. Harmon et al. 2006). Rubus hirsutus Thunberg, native of China, Japan, and the Korean peninsula (L. Lingdi and D. E. Boufford 2003), was found naturalized in a woodland in Chapel Hill, North Carolina (D. H. Goldman, pers. obs.).

(Discussion copyrighted by Flora of North America; reprinted with permission.)