Critical examinations of specimens, with literature reviews, have shown that Rubus davidianus is conspecific with R. lambertianus. Therefore, we treat R. davidianus as a new synonym within Rubus. We propose a new name, Rubus loirensis Ti R. Huang nom. nov. to replace the later homonym of R. pycnanthus Genev. Additionally, lectotypification of three names, R. davidianus Kuntze, R. malifolius Focke and R. viburnifolius Franch., are designated here after examination of previous works.

new synonyms, Rubus davidianus, R. lambertianus, R. malifolius, R. viburnifolius, species identity

Rubus L. is one of the most complicated taxonomic groups in the plant kingdom and is distributed worldwide from the lowland tropics to the subarctic region (Thompson 1995). Intraspecific/interspecific morphology and ploidy variability, apomictic tendencies and the capability of many species to hybridise widely across multiple ploidy levels, complicate Rubus taxonomy (Bammi and Olmo 1966; Alice et al. 2001; Mimura et al. 2014; Carter et al. 2019). In response to this, taxonomists disagree broadly about the number of species in the genus, with different estimates ranging from 250 (Mabberley 2017), 700 (Robertson 1974; Lu and Boufford 2003), 750 (Lu 1985), 600–800 (Thompson 1995) to more than 1000 (Jennings 1988). The most recent global taxonomic treatment of this genus was conducted by Focke in 1910, 1911 and 1914 and 12 subgenera were defined. Phylogenetic results over the past 25 years suggest that Focke’s subdivisions of Rubus are not monophyletic and large-scale taxonomic revisions are necessary. While working on the infrageneric re-classification of Rubus, we found that the taxonomic status of R. davidianus Kuntze and R. viburnifolius Franch. should be verified, especially in China (Lu and Boufford 2003). The names related to these two species, Batidaea viburnifolia Greene, R. pycnanthus Genev. and R. viburnifolius Focke, were also checked.

Rubus malifolius Focke (1890) was published, based on the collection from Chienshih, Hubei, China, A. Henry, 1885, no. 5794 (Syntypes BM000622260!; GH00040667!; K000737665!; US00097945!; Fig. 1A–D). Its critical characteristics were described as “Shrubs scandent. Leaves simple, elliptic or oblong-elliptic, base subrounded, margin inconspicuously shallowly serrate, apex acuminate, rarely acute, abaxial surface of leaves tomentose. Inflorescences terminal, racemes, bracts caducous, linear-oblong, pubescent initially, apex acute to shortly acuminate. Calyx abaxially densely tomentose-villous; sepals ovate to triangular-ovate or lanceolate. Petals white or white with pink spots, round, both surfaces thinly pubescent, base shortly clawed. Stamens many, slightly villous, anthers hirtose. Pistils much longer than stamens, ovary glabrous, styles glabrous, apex clavate. Aggregate fruit purplish-black at maturity, compressed globose, glabrous”.

Figure 1. 

A–D Syntypes of R. malifolius Focke. Specimen barcodes: BM000622260, GH00040667, K000737665 and US00097945, respectively.

Léveillé and Vaniot (1904) described R. arbor H. Lév. & Vaniot, based on the collection from Kouy-Tchéou Siao-tchang, Pin-fa, China, J. Cavaleri, May 1903, no. 1003 (Holotype E00010623!; Isotypes A00040529 (fragment with image of E00010623)!; E00313554!; K000737664!; Fig. 2A–D). Pax and Hoffmann (1922) described R. limprichtii Pax & K. Hoffm., based on the collection from Yatschou fu, Taldes Ya ho oberhalb Tschu schi ping, Hänge des Passes Tsiu gang schan, China, H.W. Limpricht, Jun 1914, no. 1564 (Syntype A00040666!; Fig. 3A). These two species were treated as synonyms of R. malifolius by Lu and Boufford (2003: 274) in Flora of China.

Figure 2. 

A Holotype of R. arbor H. Lév. & Vaniot B–D Isotypes of R. arbor H. Lév. & Vaniot. Specimen barcodes: E00010623, A00040529, E00313554 and K000737664, respectively.

Figure 3. 

A Syntype of R. limprichtii Pax & K. Hoffm. B–C Syntypes of R. viburnifolius Franch. D Holotype of B. viburnifolia Greene. Specimen barcodes: A00040666, P00746144, P00746145 and US01106201, respectively.

Franchet (1895) described R. viburnifolius Franch. in Bull. Mus. Hist. Nat. (Paris), based on the collection from les bois à Tchen-fong-chan. Yunnan, China, J.M. Delavay, Sep 1894, s. n. (Syntypes P00746144!, P00746145!; Fig. 3B–C). Greene (1906: 242) described Batidea viburnifolia Greene, based on the collection from Selkirk Mountains, US, C.H. Shaw, Aug 1904, no. 472 (Holotype US01106201!; Isotypes MIN1002232!; NY00418578!; S-G-8589 (fragment with image of NY00418578)!; Figs 3D and 4A–C). Then P.A. Rydberg (1913) merged the species into Rubus and proposed a new combination, R. viburnifolius (Greene) Rydb. However, it is a later homonym of R. viburnifolius Franch. and is, therefore, illegitimate under Art. 53.1 (Turland et al. 2018). In view of this, Berger (1925) treated it as a variety of R. idaeus L., which was named as R. idaeus var. viburnifolius (Greene) A.Berger. Focke (1910) described R. viburnifolius Focke, based on the collection from Szemao, Yunnan, China, A. Henry, no. 11714, 11714A & B and 11714C (Holotype B101154586!; Isotypes A00040762!, A00132848!, A00132850!, A00132854!; MO-255250!; Figs 4D, 5A–D and 6A). Later, Focke rejected this R. viburnifolius (1910: 117; non-Franchet 1895, non-Rydberg 1913) and replaced it with R. evadens Focke (Isotypes E00010593!, E00317755!, E00317756!; IBSC0004402!; K000737732!, K000737733!, K000737734!; US00095499!, US00996968!; Figs 6B–D, 7A–D and 8A–B). Although R. viburnifolius Franch. is the legitimate name amongst these three names, the identification of their taxonomic status is still necessary.

Figure 4. 

A–C Isotypes of B. viburnifolia Greene D Holotype of R. viburnifolius Focke. Specimen barcodes: MIN1002232, NY00418578, S-G-8589 and B101154586, respectively.

Figure 5. 

A–D Isotypes of R. viburnifolius Focke. Specimen barcodes: A00040762, A00132848, A00132850 and A00132854, respectively.

Figure 6. 

A Isotype of R. viburnifolius Focke B–D Isotypes of R. evadens Focke. Specimen barcodes: MO255250, E00010593, E00317755 and E00317756, respectively.

Figure 7. 

A–C Isotypes of R. evadens Focke. D Isotype of R. viburnifolius Focke. Specimen barcodes: IBSC0004402, K000737732, K000737733 (lower part of C), K000737734 (top part of C) and PE00020807, respectively.

Rubus lambertianus Ser. (1825) was published, based on the collection from China, Staunton, G. L., s. n. (Holotype G00316024!; Fig. 8C). Critical characteristics of the species were described as “Branchlets terete, thinly pubescent or subglabrous, with sparse, curved minute prickles. Leaves simple, cordate, base cordate, margin distinctly 3–5 lobed or undulate, serrulate. Stipules caducous, free. Inflorescences terminal usually cymose panicles, rachis and pedicels thinly pubescent, subglabrous, or glabrous. Calyx abaxially thinly pubescent, sepals ovate-lanceolate or triangular-lanceolate, margin entire, apex acuminate, margin of inner sepals grey tomentose. Petals white, obovate, glabrous, slightly shorter than or nearly as long as sepals, base clawed. Stamens many, somewhat shorter than petals; filaments broad, complanate. Pistils slightly shorter than or ca. as long as stamens, glabrous. Aggregate fruit red at maturity, subglobose, glabrous, with many drupelets, pyrenes small, prominently rugose”.

Figure 8. 

A–B Isotypes of R. evadens Focke; C Holotype of R. lambertianus Ser. D Holotype of R. pycnanthus Focke. Specimen barcodes: US00095499, US00996968, G00316024 and C10018009, respectively.

Hance (1882) described R. ochlanthus Hance, based on the collection from ad pagum Sai-ngau, secus fl. Lien-chau, Cantonensis, China, B.C. Henry, Oct 1881, no. 22021 (Holotype BM000885437!; Fig. 8D). According to the protologue, it was closely allied to R. paniculatus Sm., but was entirely distinct by the want of coloured indumentum, the much denser and more copious-flowered inflorescence and the smaller flowers. Focke (1874) described R. pycnanthus Focke, based on the collection from China, Duus, no.11 (Holotype C10018009!; Fig. 9A). In the protologue, Focke stated that R. lambertianus was different from R. pycnanthus by its lanceolate-acuminate sepals. However, these two species were treated as synonyms of R. lambertianus by Lu and Boufford (2003) in Flora of China.

Figure 9. 

A Holotype of R. ochlanthus Hance C–D Syntypes of R. davidianus Kuntze. Specimen barcodes: BM000885437, P00755281, P00755282 and P00755283, respectively.

Kuntze (1879) described R. davidianus Kuntze, based on the collection from Moupin, Su-Tchuen, China, A. David, Aug 1869, s. n. (Syntypes P00755281!, P00755282!, P00755283!; Fig. 9B–D). In Flora of China, Lu and Boufford (2003) considered it as a synonym of R. crataegifolius Bunge. However, its characters of terminal cymose panicles or axillary subracemes are different from those of R. crataegifolius. Thus, the taxonomic status of R. davidianus needs further research and its taxonomic treatment remains ambiguous.

We critically examined herbarium specimens of each species above, including all kinds of type specimens in A, BM, C, E, G, GH, IBS, K, MIN, MO, NY, P and US and checked them with protologues of each species.

The examination of herbarium specimens, identified as R. arbor, R. limprichtii and R. malifolius, indicated that they represented one species. According to Art. 11.4 of the “International Code of Nomenclature for Algae, Fungi and Plants (Shenzhen Code)” (Turland et al. 2018), R. malifolius Focke is the correct name of this species. In the same way, both R. viburnifolius (Greene) Rydb. and R. viburnifolius Focke are two later homonyms of R. viburnifolius Franch. and are, therefore, illegitimate under Art. 53.1 (Turland et al. 2018). Amongst these names with epithets such as “viburnifolius” and R. viburnifolius (Greene) Rydb., once described as Batidea viburnifolia Greene, this is characterised by “Leaves imparipinnate, 3–5-foliolate, terminal leaflet prominently petiolulate, petioles bristly, glandular-hispid and puberulent; abaxial surface of leaflets densely tomentose; stipules and bracts linear; terminal inflorescences short racemes, rarely several flower clusters in leaf axils; abaxial surface of calyx ± with needle-like prickles; pedicels densely glandular-hispid and somewhat bristly; petals white; fruit hemispherical, broad, red or yellowish, drupelets very numerous, comparatively small, falling together from the dry receptacle, pubescent”. These characters indicate that it is very closely related to R. idaeus and the differences are that the petiole, pedicel and abaxial surface of the calyx of R. idaeus have no glandular hairs. Thus, A. Berger (1925: 51) proposed R. idaeus var. viburnifolius (Greene) A.Berger. as its correct name. On the other hand, Focke (1910: 75) proposed the name R. viburnifolius Focke to represent one Chinese Rubus species, but later he noticed his error and replaced it with R. evadens Focke (1910: 117). As the specimens of R. viburnifolius Franch. which is described from China, were unable to be viewed, this was treated as a suspicious species in Flora of China (Lu and Boufford 2003). We identified two specimens of R. viburnifolius Franch. in P and describe the characters of them as: “leaves simple, elliptic or oblong-elliptic, coarsely sharply serrate, base subrounded; stipules caducous, linear-oblong to ovate-lanceolate; terminal inflorescences racemes; rachis and pedicel densely tomentose-villous, gradually glabrescent, finally glabrous; bracts caducous, linear-oblong, apex acute to shortly acuminate; calyx abaxially densely tomentose-villous, sepals entire”. All of these characters indicate that R. viburnifolius Franch. is conspecific with R. malifolius and, therefore, R. viburnifolius Franch. is a later synonym of R. malifolius.

The examination of herbarium specimens, identified as R. lambertianus, R. ochlanthus and R. pycnanthus Focke, indicated that they represent the same species and, therefore, R. lambertianus is the correct name of this species. R. davidianus is a Chinese Rubus species described by Kuntze and three specimens of it have been identified in P, characters of them being described as: “shrubs; leaves simple, broadly ovate, rarely oblong-ovate, abaxially pilose, more densely so along veins, rarely glabrous, with sparse, minute prickles along mid-vein, adaxially pilose or hairy only along veins, cordate at base, margin distinctly 3–5-lobed or undulate, serrulate, apex acuminate; stipules and bracts narrower, less than 2 × 1 cm, linearly lobed; terminal inflorescences cymose panicles, axillary ones often subracemes, shorter, sometimes flowers few in clusters in leaf axils; pedicel 0.5–1 cm long; calyx abaxially thinly pubescent, sepals ovate-lanceolate or triangular-lanceolate, undivided; petals obovate, glabrous, slightly shorter than or nearly as long as sepals”. These characteristics are consistent with those of R. lambertianus, which indicate that R. davidianus is a later synonym of R. lambertianus, not R. crataegifolius.

In the process of R. pycnanthus Focke identification, we found that another plant, occurring in Haute-Vienne, Saint-Sulpice-les-Feuilles, Thias, Lamy, Angers, Maine-et-Loire, France, was also named as R. pycnanthus Genev. (Genevier 1880). Actually, Genevier (1868) firstly published it as R. pyramidatus Genev. Then he rejected it because Müller (1859) had published a name with the same epithet “pyramidatus” for a German plant. According to Art. 53.1 (Turland et al. 2018), R. pycnanthus Genev. is also a later homonym of R. pycnanthus Focke and, therefore, a new name, R. loirensis Ti R. Huang nom. nov., is proposed.

Lu and Boufford (2003: 285) listed R. viburnifolius Franch. at the end of Flora of China to indicate that it had been described from Yunnan, China, but they have not seen any specimens and are, therefore, unable to treat it. Meanwhile, they stated that further revision of this species was necessary. In this paper, we carried out critical examinations of herbarium specimens, from which morphological characters of R. arbor, R. limprichtii, R. malifolius and R. viburnifolius were studied. Morphological characters of R. arbor, R. limprichtii, R. malifolius and R. viburnifolius Franch. indicate that they represent the same species and, therefore, R. malifolius is the correct name according to the “International Code of Nomenclature for Algae, Fungi and Plants (Shenzhen Code)” (Turland et al. 2018). Additionally, both R. viburnifolius (Greene) Rydb. and R. viburnifolius Focke are later homonyms of R. viburnifolius Franch., in which R. viburnifolius (Greene) Rydb. was replaced by R. idaeus var. viburnifolius and R. viburnifolius Focke was replaced by R. evadens.

Rubus davidianus was treated as a synonym of R. crataegifolius Bunge by Lu and Boufford (2003: 236) in Flora of China. However, examination of herbarium specimens indicates that there are distinct differences between the two species. The differences are: the former has inflorescences with terminal cymose panicles, axillary ones often subracemes, shorter, sometimes flowers few in clusters in leaf axils; stipules and bracts narrower, less than 2 × 1 cm, linearly lobed; the latter has inflorescences terminal, rarely axillary, short racemes or flowers several in cluster; stipules and bracts linear, entire. Thus, we conclude that R. davidianus and R. crataegifolius should represent two different species of Rubus and R. davidianus should be a synonym of R. lambertianus. Three specimens stored under R. pyramidatus P.J. Müll. in P were found; however, the specimens stored under either R. pycnanthus Genev. or R. pyramidatus Genev. could not be traced. Based on the existing characters of R. pyramidatus P.J. Müll. and R. pycnanthus Genev., we can identify that these two species are different from that of R. pycnanthus Focke. Though there are old attempts to synonymise R. pycnanthus Genev., the taxonomic status of R. pycnanthus Genev. should still be studied.

Species identification of Rubus species indicates that many homonyms and synonyms still exist in the genus Rubus, especially when they were more common in the 18^th, 19^th and 20^th century. This could be interpreted in three ways. First, because of the propensity for interspecific hybridisation, polyploidy and apomixis, morphological characters of the species under this genus are highly variable and diverse. This makes species division and identification very difficult. Second, the original publications of species are often kept in the libraries of various scientific research institutions and some original publications are even kept in private collections. Objectively, this increases the difficulty for people to obtain and read the information of species publications. Third, examination of type specimens could not be easily accessed since digitisation of specimens was not yet widespread. Therefore, species names of Rubus, once not given sufficient attention or had not been discovered, should be emphasised in further taxonomic studies, using the integrative morphological characters and integrative systematics.

The authors would like to express gratitude to curators of the herbaria: A, BM, C, E, G, GH, IBSC, K, LE, MIN, MO, NY, P, PE, S-G and US. We are grateful to an anonymous reviewer for valuable comments. This research was financed by the Beijing Municipal Education Commission under Grant KM201910020016.