All the sequences in this study were retrieved from GenBank and the accession numbers of the newly-determined sequences in this study are shown in Suppl. material 1.

A total of 136 COI gene sequences of 1448 bp in length of two Sinobdella species were here amplified. Forty haplotypes (including four outgroups) were detected from the sequences (Suppl. material 1). These sequences were used for molecular phylogenetic analysis along with four sequences from the outgroups: Monopterus albus (one), Mastacembelus armatus (one), M. favus (one) and M. erythrotaenia (one). The sequences consisted of 969 conserved sites, 479 variable sites, 327 parsimony informative sites and 152 singleton sites. The nucleotide frequencies were 25.9% (A), 30.8% (T), 26.0% (C) and 17.2% (G).

Given that Bayesian Inference (BI) and Maximum Likelihood (ML) analyses produced overall identical topologies, only the BI tree with Bayesian posterior probabilities (PP) and bootstrap support (BS) value were provided in Fig. 7. From the tree topologies, samples of S. longitubulus were strongly supported (PP = 1.0/BS = 100) to cluster into a lineage further constituting a well-supported (PP = 1.0/BS = 100) clade together with S. sinensis.

Figure 7.  

Bayesian Inference tree based on the COI gene for two species of Sinobdella and three species of Mastacembelus. Bayesian posterior probabilities (> 0.5), maximum Likelihood bootstrap values (> 50%) are shown, respectively. Dashes represent nodes with bootstrap support lower than 50%.

The genetic distances (p-distances) within and between species were calculated. Intraspecific genetic distance for sampled species of S. sinensis and S. longitubulus were 0.5% and 0.1%, respectively. Interspecific genetic distance between both were 10.5%.

Bleeker (1870a) described Rhynchobdella sinensis, based on a single 199 mm TL specimen caught in China. Despite no indication of the accurate type locality in the original description, Bleeker (1870b) noted that this species was one of 44 species then identified from Chinese specimens predominantly collected from the Yang-Tse-Kiang (or Yangtze River = Chang-Jiang in mainland China) and associated Lake Po Yang (= Lake Poyang) and Kan-Kiang River (presently Gan-Jiang, a stream of Lake Poyang) and the waters of Ning-Po (presently Ningbo). He further stated that these Chinese specimens, received from the administration of Jardin royal des plantes médicinales (the predecessor of the Muséum national d'Histoire naturelle, MNHN), constituted part of the collections by C.P. Dabry de Thiersant, G.E. Simon and A. David intended for the Museum in Paris (MNHN); the board of the Museum in Paris invited him to identify them. Apparently, the type specimen of S. sinensis ought to be housed at MNHN rather than at BMNH (British Museum of Natural History) and the collector of the type is likely one of the three persons mentioned.

There are seven records of Chinese spiny eels currently housed in MNHN (Table 2). Seven specimens [MNHN-IC-0000-5230 (two), 5573 (four) and 7490 (one)] of unknown source were collected from China by A. David in 1869 and 1870, respectively. Four specimens [MNHN-IC-0000-5021 and 5022 (two), from the Yangtze River (or today’s Chang-Jiang); 7851 (one, from mountains of Jiangxi Province); and 7358 (one, from China, but without precise locality) were respectively collected by C. P. Dabry de Thiersant in 1863, 1868 and 1873. A comparison of X-ray photographs of two Chinese spiny eels reveals that Sinobdella sinensis has more vertebrae between the first and second anal spine pterygiophore than Mastacembelus armatus (four or five vs. one or two) (Fig. 8). In terms of our examination on the X-ray photographs available in MNHN, the first and second pterygiophore of anal spines are separated by one or two vertebrae in two specimens (MNHN-IC-0000-5021 and 5022) (Fig. 9b, c), so indicating their misidentification of M. armatus. The specimen (MNHN-IC-0000-7490) is not the type of S. sinensis as it was collected in 1870, the same year as the publication of its original account, but slightly after the collection time (1868) of the type specimen of this species (see explanation below). The possibility is ruled out that the specimen (MNHN-IC-0000-7358) is the type of S. sinensis as it was caught in 1873, well after the publication of its original description. It is probable that the type of this species is amongst those caught by A. David (MNHN-IC-0000-5230 and 5573) and C. P. Dabry de Thiersant (MNHN-IC-0000-7851) in 1868. G.E. Simon is not the collector as spiny eels are not found in Chinese fish specimens, available in the MNHN collection, caught by him.

a

b

Figure 8.

X-ray photographs of lateral body for Mastacembelus armatus and Sinobdella sinensis. The red frame indicates the differences that exist between the anal spines of the two species.

a: Mastacembelus armatus, IHB 202010051228, caught from Li-Jiang of Xi-Jiang at Yangshuo County;  
b: Sinobdella sinensis, IHB 202010051336, caught from Luoqing-Jiang of Xi-Jiang at Yongfu County, Guangxi Province.  

a

b

c

Figure 9.

X-ray photographs of lateral body of MNHN collection specimens.

a: Sinobdella sinensis (MNHN-IC-0000-7851);  
b: Mastacembelus armatus (MNHN-IC-0000-5021);  
c: Mastacembelus armatus (MNHN-IC-0000-5022).  

During 1868 to 1870, A. David made the second trip to central and western China for specimen collection (Luo 2005). His collection locations were in the Chang-Jiang Basin, including Zhenjiang and Jiangsu Province, Kiu-Kiang (or Jiujiang) and Lake Po Yang (or Poyang) of Kiang-Si (or Jiangxi) Province, Shanghai, Hankou and Yichang of Hubei Province, Chongqing and Chengdu of Sichuan Province; all collected specimens were sent to MNHN. Undoubtedly, six specimens of Chinese spiny eel (MNHN-IC-0000-5230 and 5573) were caught by A. David in the Chang-Jiang Basin. Information available in MNHN indicates that C.P. Dabry de Thiersant made a fish collection in the Chang-Jiang and mountain rivers of Jiangxi Province during 1868 to 1869. Based on fish specimens collected by him, Bleeker subsequently described many new Chinese species (Bleeker 1870a, Bleeker 1870b, Bleeker 1870c, Bleeker 1870d, Bleeker 1871, Bleeker 1872, Bleeker 1874). It cannot exclude the possibility that the specimen (MNHN-IC-0000-7851), collected in 1868 by Dabry de Thiersant from mountainous rivers of Jiangxi Province, was used as the type by Bleeker (1870a) to describe Rhynchobdella sinensis (now Sinobdella sinensis). No matter who is the collector, it is for certain that the type of this spiny eel is from the Chang-Jiang Basin. The specimen (MNHN-IC-0000-7851) from the mountain streams of Jiangxi Province is most likely the type as it agrees with the original description in, amongst others, body size.

Chinese spiny eels under the name of S. sinensis in the BMNH collection are also amongst mastacemblids examined by both Travers (1984a), Travers (1984b) and Vreven (2005). Three specimens (BMNH 1888.3.23: 60-62) of 119-169 mm TL and one specimen (BMNH 1895.5.31: 13-14) of 202 mm TL, which were donated by F. W. Styan, according to information available in this Museum, are from Kiu-Kiang (= Jiujiang City), Jiangxi Province (in the lower Chang-Jiang Basin) and in Shanghai (in the estuary of this river), respectively. Although three specimens BMNH 1888.3.23:60-62 were listed as types (Travers 1984a) or syntypes (Vreven 2005), they are unlikely the types of S. sinensis as they disagree with its original description in body size, far less than 199 mm TL of the type, as well as the collection time as indicated by their catalogued number, well after the publication of its original description. The specimen BMNH 1895.5.31:13-14 was listed as a syntype with a question mark (Vreven 2005), possibly due to its similar body size, close to 199 mm TL given in the original description for the type. It is not the type of this species because of its later collection time, as indicated by its catalogued number.

The discovery of this new species from the middle Zhu-Jiang Basin in Guangxi Province necessitates a re-diagnosis of S. sinensis and a delineation of its distribution in China. Characters typical for S. sinensis are given in the diagnosis. According to these characters, this species is extensively known from the mid-lower Chang-Jiang Basin. Populations of river basins north of this river, such as the Huai-He, Hai-He and Luan-He are conspecific with the species, as evidenced by molecular data (Fig. 7). This spiny eel is also found in the Li-Jiang and Luoqing-Jiang, two stream tributaries of the middle Xi-Jiang of the Zhu-Jiang Basin. The plausible explanation for this is that the spiny eel possibly dispersed into the two rivers from the Xiang-Jiang via the Ling Canal built in 214 BC.

Except for the count of vertebrae (Table 3), no distinct morphological and genetic variations are found between populations of the middle Zhu-Jiang and mid-lower Chang-Jiang Basin for S. sinensis. Sixteen specimens examined from Lingchuan County and Guilin City (in the Li-Jiang) and Yongfu County (in the Luoqing-Jiang) had 76-77 (mean 76.42) vertebrae, slightly fewer than 77-79 (mean 78) of four specimens examined from nearby Xing’an County (in the Xiang-Jiang, Guangxi Province; 76-78 (mean 77.6) of ten specimens from the Xiang-Jiang and Yuan-Jiang, two streams of Lake Dongting in Hengdong and Mayang Counties, Hunan Province; and 76-81 (mean 79.11) of 21 specimens collected from the Han-Jiang, the largest tributary of the Chang-Jiang Basin, in Xiangyang City and She-Shui in Dawu County, Fu-He in Sui County and Huan-He in Xiaochang County, three streams on the northern bank of the middle Chang-Jiang mainstream in Hubei Province. This variation can be plausibly explained by Jordan’s rule that stipulates a geographical tendency for populations of the same fish species from higher latitudes to have an increased number of vertebral centra (Jordan 1891, McDowall 2007).

This phenomenon is also observed for S. sinensis in east China. For example, four specimens caught from the Qiantang-Jiang Basin in Changshan County, Zhejiang Province, possessed 76-77 (mean 76.3), fewer than 77-79 (mean 77.8) of nine specimens from the Gan-Jiang (in Poyang Lake system) in Wanzai and Yongxin Counties, Jiangxi Province and 78-81 (mean 79.9) of seven specimens from the lower Chang-Jiang Basin in Susong and Tongcheng Counties and 77-79 (mean 77.6) of five specimens from the Huai-He Basin in Shangcai County of Henan Province and Huoshan County of Anhui Province. It seems that marked variations in the mean number of vertebrae were found between sampling locations north and south of the latitude (30°N). The highest number of vertebrae, though, was not found in the northern-most sampling location in the Luan-He (Luanzhou City). Mean vertebrae count also showed a longitudinally increasing variation from the Qiantang-Jiang to the middle and lower Chang-Jiang Basin. Collectively, the number of vertebrae for S. sinensis varies geographically, probably in association with temperature amongst different regions.

Mastacembelus dienbienensis is a Vietnamese species firstly described by Nguyen (2005) from the Nam Nua system in Dien Bien Phu City, in the Mekong River Basin of Vietnam. Although Kottelat (2013) ascribed it to Sinobdella, this generic assignment remains indeterminacy as this genus is currently unknown from the Mekong River Basin. Endruweit (2014) retained this spiny eel in Mastacembelus and confirmed that its four type specimens have been lost in RIA 1(Research Institute for Aquaculture No. 1, Bac Ninh, Hanoi). For this reason, its validity and generic placement remain yet to be resolved, only if topotypical specimens become available. For the time being, this species is excluded from Sinobdella.

Mastacembelus kobayashii was firstly described by Oshima (1926) from Taiwan Island. This spiny eel is only known by two type specimens in the original description. No additional specimens have been collected so far; its validity has remained yet to be resolved. Taiwanese spiny eels were previously identified as Macrognathus aculeatum (Bloch, 1786) (=Sinobdella sinensis) (Tzeng 1986). It has been considered as a subspecies of S. sinensis in the annotated checklist and type catalogue of fish genera and species described from Taiwan by Ho and Shao (2011). This taxonomic treatment is followed in this study.