Handheld dip nets were utilized as the primary tool for sample collection. According to documented literature, the distribution range of the B. pingi species group includes Lingyun County, Tian’e County, Tianlin County, and Longzhou County in Guangxi Province, and Guangnan County in Yunnan Province, where specimen collection was conducted (Fang 1930; Chen 1990; Chen and Tang 2000). Due to various natural and anthropogenic factors leading to habitat alterations, no specimens were collected in Tian’e County and Longzhou County. Extensive exploration of other potential distribution areas was conducted, leading to the discovery and collection of specimens in the Youjiang District of Baise City, Debao County, Xilin County, Napo County, and Wuming District of Nanning City in Guangxi Province, and Wenshan Zhuang and Miao Autonomous Prefecture in Yunnan Province. Additional collections of B. szechuanensis (Fang, 1930), B. liui Chang, 1944, B. niulanensis Chen, Huang & Yang, 2009, B. polylepis Chen, 1982, B. kweichowensis (Fang, 1931), and B. leveretti (Nichols & Pope, 1927) specimens were conducted from the middle and upper reaches of the Yangtze River in Sichuan Province, the Jinsha River basin and Nanpan River basin in Yunnan Province, the Li River basin in Guangxi Province, and the Duliu River basin in Guizhou Province, as well as from Hainan Province for comparative purposes. Following euthanasia, specimens were preserved in a 95% ethanol or 0.3% formalin solution, with 3–5 individuals randomly selected from each batch for mitochondrial cytb gene sequencing. Despite repeated attempts, no specimens of B. yunnanensis, B. huangguoshuensis Zheng & Zhang, 1987, B. intermedia Tang & Wang, 1997, and B. cyclica Chen, 1980 were obtained, and morphological data were cited from their original descriptions based on examination of their holotypes or paratypes. All fish collections followed the Law of the People’s Republic of China on the Protection of Wildlife, and we strictly adhered to all applicable international, national, and institutional guidelines for the care and use of animals.

For other questionable species with unobtainable specimens and disputed taxonomic status, their original morphological descriptions were referenced.

Countable characteristics were examined using a stereoscopic microscope. The last branched anal-fin and dorsal-fin rays located on the last complex pterygiophore were counted as a single ray (Fig. 1). A considerable percentage of the population belonged to a phenotype characterized by attenuated and unbranched terminal fin rays in the paired fins. This specific anatomical feature was quantitatively assessed as equivalent to 0.5 of a standard branched fin ray. Following the photographic documentation of specimens in dorsal, lateral, and ventral views, measurements were conducted using TPSDIG2 software (Rohlf 2016), as depicted in Fig. 1. Statistical analyses of the measurable characteristics were conducted using SPSS v.26.0 (IBM Corp. 2020) to assess the significance of differences.

Figure 1. 

A. Morphometric measurement methodology for species within the genus Beaufortia: 1 - Standard length (SL); 2 - Eye diameter (ED); 3 - Gill slit length (GSL); 4 - Pectoral fin base length (PBL); 5 - Head length (HL); 6 - Body depth (BD); 7 - Dorsal fin base length (DBL); 8 - Anal fin base length (ABL); 9 - Post-anal fin length (PoAL); 10 - Caudal peduncle length (CL); 11 - Caudal peduncle depth (CD); 12 - Head length (HL); 13 - Snout Length (SnL); 14 – Head width (HW); 15 - Interorbital distance (IOD); 16 - Distance between nostrils (ND); 17 - Pre-dorsal Length (PrDL); 18 - Pectoral fin disc length (PL); 19 - Head width at pectoral fin origin (HWPO); 20 - Mouth slit width (MW); 21 - Pectoral fin disc width (PW); 22 - Body width (BW); 23 - Pelvic fin disc width (PvW); 24 – Pre-pectoral length (PrPL); 25 – Pre-pelvic length (PrPvL); 26 - Pelvic fin disc length (PvL); 27 - Pre-anal pore length (PrApL); 28 - Pre-anal fin length (PrAL); B. The last branched anal-fin and dorsal-fin rays located on the last complex pterygiophore were counted as a single ray.

DNA was extracted from the right pectoral fin rays using a kit produced by Sangon Biotech (Shanghai) Co., Ltd., following the protocol provided. Primers GAC TTG AAG AAC CAC CGT TGT TAT T 5’–3’ and TCT TCG GAT TAC AAG ACC GAT GCT TT 5’–3’ were employed to amplify the cytb gene. The PCR mixture consisted of 1µL DNA template, 1µL of each primer (10 µM), 12.5µL Taq Mix (Sangon Biotech), and 9.5µL deionized water, totaling a reaction volume of 25µL. The PCR conditions included an initial denaturation at 95 °C for 3 minutes, followed by 35 cycles of 94 °C for 30 seconds, 55 °C for 45 seconds, and 72 °C for 60 seconds, with a final extension at 72 °C for 5 minutes and a hold at 4 °C. Amplified products were sequenced by Sangon Biotech, and the sequences were verified for accuracy before submission to the National Center for Biotechnology Information (NCBI). The collection sites for the molecular materials are listed in Table 1.

A total of 35 mitochondrial cytb gene haplotypes were identified. The cytb sequence of Plesiomyzon baotingensis (KF732713), retrieved from NCBI, was utilized as the outgroup, with a haplotype of B. szechuanensis (NC027291) included as additional material. Phylogenetic analyses were conducted using PHYLOSUITE (Zhang et al. 2020) with sequence alignment performed by MAFFT (Katoh and Standley 2013) set to automatic strategy and normal alignment mode. ModelFinder (Kalyaanamoorthy et al. 2017) was used to determine the best-fit models for both Maximum Likelihood (ML) and Bayesian Inference (BI) methods. The ML phylogenetic tree was constructed using IQ-TREE (Nguyen et al. 2015), with the best-fit model according to BIC being MGK+G4, and the outgroup set to Plesiomyzon baotingensis, with default parameters for other settings. The BI phylogenetic tree was constructed using MRBAYES 3.2.6 (Ronquist et al. 2012), with the model set to HKY+F+G4 and the outgroup as Plesiomyzon baotingensis, maintaining default parameters for other settings. The resulting tree files were visualized and embellished using FIGTREE v1.4.4 (Rambaut 2018).

Genetic distances were calculated using MEGA11 (Tamura et al. 2021) with the Kimura 2-parameter model (rates among sites: G4), computing distances among all samples, average interspecific genetic distances, and average intraspecific genetic distances.

A total of 35 mitochondrial cytb gene haplotypes were obtained. The computed genetic distances revealed that B. granulopinna sp. nov. shared the smallest average genetic distance with B. zebroida at 10.80% (with the maximum interspecific genetic distance reaching 11.37%). Intraspecific genetic distances ranged from 0.26% to 5.12%, with an average of 3.63%, surpassing the average intraspecific genetic distances of all other species within the same genus. B. viridis sp. nov. exhibited a minimum average interspecific genetic distance with B. pingi of 4.60% (with a maximum interspecific genetic distance of 5.2%), significantly exceeding the current minimum intrageneric interspecific distance. Intraspecific genetic distances ranged from 0.09% to 2.80%, with an average of 1.47%, exceeding the average intraspecific genetic distances of all other congeneric species except for B. granulopinna sp. nov. B. zebroida had a minimum average genetic distance with B. pingi of 5.62% (with the maximum interspecific genetic distance at 5.88%), considerably above the current minimum intrageneric interspecific distance. The genetic distances between the aforementioned species and those within the B. pingi species group were smaller, indicating their affiliation with the B. pingi species group, aligning with morphological classification results.

Notably, the genetic distance between B. niulanensis and B. szechuanensis was only 0.53%, whereas the B. cf. szechuanensis collected from Qijiang, Chongqing (a tributary on the southern bank of the upper Yangtze River) morphologically conformed to B. szechuanensis but exhibited a genetic distance of 2.30% from B. szechuanensis, exceeding the genetic distance between B. polylepis and B. szechuanensis (2.20%) (see Table 5).

Phylogenetic trees generated using maximum likelihood and Bayesian inference methods exhibited identical topological structures, with all nodes demonstrating high support values. All species within the B. pingi species group coalesced to form a single clade, with each species within this clade establishing distinct lineages. Among these, B. pingi and B. viridis sp. nov. were identified as sister taxa positioned at the derived end of the phylogenetic tree, followed by B. zebroida as a sister group to the aforementioned species. B. granulopinna sp. nov. was situated at the most basal position within the B. pingi species group clade, with a relatively longer branch length indicating an early divergence.

The remaining species formed another clade, distinctly divergent from the B. pingi species group, which aligns with the classification results based on morphological characteristics. Within this clade, two significantly divergent lineages were identified. One lineage was composed of all haplotypes of B. kweichowensis and B. leveretti. The other lineages included B. szechuanensis, B. polylepis, B. niulanensis, and B. cf. szechuanensis. (see Fig. 16).

Figure 16. 

Phylogenetic tree of Beaufortia species based on haplotypes of mitochondrial cytb genes, with the tree topology inferred through maximum likelihood (ML) and Bayesian inference (BI) methods. Support values for each branch are indicated at the nodes (ML/BI), with “*” denoting bootstrap values of 100% or posterior probabilities of 1, and “-” indicating insufficient genetic variation to display posterior probabilities.

The preceding phylogenetic results have confirmed that the Beaufortia pingi species group constitutes a monophyletic clade and exhibits significant genetic divergence from other Beaufortia species. Morphologically, this group is distinctly differentiated from other species. The lower lip of the species group can be represented by a pinnate-type appearance, being overall slender with a central indentation and multiple lobed sides. There is variation among individuals in the number of lobes and the depth of the clefts, with most specimens possessing no fewer than three lobes and the depth of clefts exceeding that found in species outside this group. The extremities of the lower lip extend to the corners of the mouth, with several wart-like protrusions present on the extended parts (Fig. 6). In contrast, species outside this group possess a dicot-type lower lip, which is broad and short with a central indentation, and sides that are not lobed but occasionally have 0–2 shallow notches. The ends of the lower lip do not extend to the corners of the mouth, or the extension is poorly developed and lacks protrusions (Fig. 7). Additionally, the Beaufortia pingi species group is characterized by distinct vertical stripes along the body sides, a feature absent in other Beaufortia species.

B. pingi has more branched rays in its paired fins, with an average of 22.19±0.78 in pectoral fins and 18.64±0.97 in pelvic fins. In contrast, B. granulopinna sp. nov. exhibits fewer branched rays in its paired fins, averaging 20.11±0.77 in pectoral fins and 16.51±0.86 in pelvic fins.

Differences in morphometric proportions were observed among the four species within the Beaufortia pingi group. B. granulopinna sp. nov. and B. pingi have comparatively taller bodies, as reflected in lower mean SL/BD ratios (5.10±0.43 and 5.09±0.71, respectively) compared to those observed in B. viridis sp. nov. and B. zebroida (5.31±0.54 and 5.47±0.33, respectively). B. pingi is distinguished by a broader body, as evidenced by smaller mean ratios of SL/BW and SL/PW (4.2±0.36 and 7.42±0.39, respectively), a shorter head, indicated by a larger mean SL/HL (4.9±0.18), and a narrower mouth, denoted by a larger mean HW/MW (5.79±0.5). Compared to the other three species, B. granulopinna sp. nov. features a shorter and thicker caudal peduncle, as shown by the smaller mean SL/CD and PoAL/CD (8.36±0.52 and 1.29±0.11, respectively).

Variability in morphometric characteristics is observed within populations of B. viridis sp. nov. from different regions. Specimens from Lingyun County possess a more robust anterior body segment, with wider and taller heads and trunks (PrPvL/HW=1.53±0.08, PrPvL/PW=0.99±0.03, PrPvL/BW=1.73±0.07, PrPvL/BD=1.98±0.15, PrPvL/HD=2.48±0.12, PrPvL/PW=1.25±0.05), and a thicker caudal peduncle (PrPvL/CD=3.59±0.28). The population from the type locality, Wuming District, Nanning City, exhibits fewer branched rays in the paired fins (pectoral fins 20.25±0.68, pelvic fins 16.5±0.71).

Compared to the type locality, the population from Debao County is characterized by a broader and flatter body (PW/BD=2.32±0.19) and a thinner caudal part (PoAL/CD=1.69±0.09) (see Table 4).

It is noteworthy that, despite the statistically significant and visually discernible differences in measurable characteristics, the presence of overlap among different species precludes recommending these characteristics as reliable bases for species identification.

In 1930, Fang described Gastromyzon pingi with the type locality specified as Lingyun County in 1928. He mentioned two collection sites, one near the boundary of Yunnan Province (type locality) and the other in Damaoping Village. At that time, the jurisdiction of Lingyun County was more extensive, bordering Yunnan and encompassing the current eastern area of Tianlin County. The county’s watershed includes the Xiyang River, Bo′ai River, Jia River (currently Leli River), and Chengbi River. This broad and somewhat vague collection range coincides with the primary distribution area of the Beaufortia pingi species group. Given the relatively short geographical distances between distribution areas of different species within this region, it is highly probable that the type specimens could actually belong to different species. The paratype specimen NO. 955 exhibits a notably lower count of pectoral fin rays, a distinguishing characteristic of B. granulopinna sp. nov. from B. pingi. Excluding the lateral line scale count, the morphological features of the remaining specimens align with those of B. pingi specimens collected from the Chengbi River basin within the current boundaries of Lingyun County. The scales of the B. pingi species group are exceptionally fine, often less than 0.5 mm in diameter, suggesting that the original counts might have been inaccurately high due to the limitations of observation equipment.

Apart from the Chengbi River basin, B. pingi was also collected from the Jian River basin in Debao County. Compared to the holotype, these specimens have slightly fewer ventral fin rays and a somewhat broader and flatter body, but they share the main differential features of the species. The genetic distance of the mitochondrial cytb gene is less than 1%, indicating minor genetic variation, and phylogenetically, these specimens cluster within the same lineage as B. pingi.

In summary, the species within the Beaufortia pingi group are geographically proximate and morphologically similar. The type locality of B. pingi is relatively indistinct, and the age of the specimens complicates identification. The inclusion of other species among the paratypes, the original vague description, and some inaccuracies in detail present challenges. Considering these issues, we have conducted a redescription of B. pingi based on extensive specimen collection, morphological comparisons, and molecular phylogenetic analysis.

Gastromyzon pingi zebroidus was also described in the same article, in which Fang (1930) described Gastromyzon pingi with a single-type specimen collected from the Dongui River along the China-Vietnam border in Longzhou County, Guangxi Province. Since its publication, no further records of this species have been found in that basin. Despite multiple surveys in Longzhou County, the species was not rediscovered, and the local rivers, mostly altered by human activities and facing severe bioinvasions, significantly reduced the habitats suitable for gastromyzonids.

The limited information from a single specimen, coupled with the low distinctiveness of the illustrations and the relatively vague morphological description of Gastromyzon pingi zebroidus, has led to significant controversy over the identity of this species and multiple changes in its taxonomic status.

In the original description, Fang (1930) distinguished this subspecies by the anal pore’s proximity to the posterior edge of the ventral fins and the presence of stripes along the entire flank (vs. the anal position being further from the posterior edge of the ventral disc and stripes being pronounced at the caudal peduncle in Gastromyzon pingi). Later, Hora (1932) elevated its taxonomic status to species level as Beaufortia zebroidus without providing justification. Nichols (1943) followed Fang’s (1930) perspective, considering it a subspecies of Gastromyzon pingi.

Chen (1980), upon examination of two specimens from the Xijiang River (location unspecified), with total lengths of 29 mm and 52 mm, respectively, observed that both specimens exhibited pronounced or faint vertical stripes and noted the anal pore situated at a considerable distance from the ventral fins. Consequently, Chen tentatively synonymized Beaufortia zebroidus with Beaufortia pingi, pending further collection of specimens for a definitive conclusion. According to our observations, the vertical stripes on juvenile stages of B. pingi typically do not disappear, and the specimen measuring 29 mm in length is likely in a late juvenile stage. Therefore, the species in question is inferred to be B. pingi.

Yue (1981), upon examination of nine specimens from Lingyun County, Guangxi, identified them as B. pingi and noted the presence of 2–4 transverse vertical stripes on the caudal peduncle. The presence and clarity of stripes anterior to the caudal peduncle varied among individuals, with some exhibiting up to 10 distinct stripes. The anal pore was observed to be closer to the anal fins than to the posterior edge of the ventral fins, with distance varying across the specimens. Notably, one specimen, No.74664, possessed an anal pore precisely at the posterior edge of the ventral fins. Based on these findings, Yue proposed that B. zebroidus could be a synonym of B. pingi or not a valid subspecies. Zheng (1989) examined 15 specimens from Lingyun County, Guangxi, and Xingyi County, Guizhou, and identified them as B. pingi. Zheng observed the terminal rays of the ventral fins connecting medially on both sides, with a notch at the end (presumably a typographical error for “a pair of notches”), the posterior edges located near the anal pore or slightly distant, and 11–13 transverse bands across the dorsal midline accompanied by approximately 20 vertical stripes. The presence of unclear stripes, visible only at the caudal peduncle in some individuals, led Zheng to consider B. zebroidus a junior synonym of B. pingi. Both researchers examined the specimens from the type locality, and given that Lingyun County is habitat to both B. pingi and B. viridis sp. nov., the possibility of these specimens comprising two distinct species cannot be excluded.

Subsequently, Chen (1990) examined 21 specimens from the Xiyang River in Guangnan County, Yunnan, identifying them as B. pingi. The observations revealed that the tips of the pelvic fin rays were proximal to the anal pore in all specimens, with numerous dark and light stripes along the dorsal midline and flanks. Chen thus regarded B. zebroidus as a junior synonym of B. pingi. Our research indicates that the species distributed in the Xiyang River is B. viridis sp. nov., and given the consistent presence of vertical stripes in all specimens, the referenced species is likely B. viridis sp. nov.

Later, Chen and Tang (2000) measured three specimens from Tian’e, Guangxi, noting that the posterior edges of the pelvic fins were proximal to the anal pore. The dorsal and lateral sides exhibited approximately 20 black vertical stripes, with the pre-pelvic length significantly exceeding the distance from the pelvic fin origin to the anal fin origin. The branched ray count in the paired fins (pectoral 19, pelvic 15–16) was slightly lower than that observed in the specimens from the type locality of B. pingi (pectoral 21–23, pelvic 18–20), leading them to recognize B. zebroidus as a valid species. We noted that the branched ray counts in the paired fins of the specimens collected from Tian’e County fell within the variation range of the specimens from the type locality of B. viridis sp. nov., with the pre-pelvic length generally greater than the length from the pelvic origin to the anal origin (mean PrPvL/POAO=1.19±0.74), distinct from the specimens from the type locality of B. pingi (PrPvL/POAO=0.98±0.66). Hence, the specimens collected from Tian’e County are identified as B. viridis sp. nov.

Chen and Zhang (2006), after examining six specimens from Tianlin County and Lingyun County and referencing Chen and Tang (2000), encountered individuals exhibiting characteristics of both B. pingi and B. zebroidus (presence or absence of vertical stripes, count of branched rays in paired fins, PrPvL/POAO ratio), thereby considering B. zebroidus a synonym of B. pingi. Tianlin County serves as the type locality for B. granulopinna sp. nov. Our statistical analysis indicates that a certain proportion of vertical stripes disappear in the adult stages of both B. granulopinna sp. nov. and B. pingi (54.76% and 94.00%, respectively). The branched ray counts in the paired fins of B. granulopinna sp. nov. are also marginally less than those of B. pingi, and the PrPvL/POAO ratios for both species fluctuate around 100% (as shown in Table 3). Moreover, we observed that the photographs provided by the authors (original fig. X 98), which depicted B. granulopinna sp. nov., showed the first few rays of the pectoral fins exhibiting well-developed tubercles. Therefore, the batch of specimens actually represents a mix of B. pingi and B. granulopinna sp. nov. from their type localities.

Finally, Kottelat (2012) recognized B. zebroidus as a valid species, changing the suffix to -a, and synonymized B. fasciolata, B. multiocellata, and B. triocellata, collected from the China-Vietnam border, with B. zebroida.

After investigating all possible distribution areas and combining morphological analysis with mitochondrial cytb gene phylogenetic analysis, we conclude the following:

The type locality of B. zebroida is the border river between China and Vietnam. Among all the literature records, the ones closest to the type locality are the collection sites of B. fasciolata, B. multiocellata, and B. triocellata. Although the type specimens of these three species are lost, their original descriptions and unclear illustrations still confirm that they belong to the same species as our specimens collected from the southeastern Yunnan Province to the western Guangxi border area with Vietnam. Fang (1930) mentioned the specimen color as pale fleshy, which aligns with the body color of the aforementioned populations but not with B. pingi and B. viridis sp. nov. Thus, we consider B. zebroida a species distributed from Maguan County in Yunnan Province to Longzhou County in Guangxi Province along the China-Vietnam border, as well as in northeastern Vietnam. The region contains small tributaries flowing to both the Red River and the Zuo River (Pearl River system), closely spaced due to the karst landscape, facilitating stream capture and exchange. The type locality, Longzhou County, represents the eastern boundary of the species’ distribution, but due to habitat destruction and bioinvasions, no further records of this species have been found there. Besides this area, the species has been discovered in Maguan County, Guangnan County, Xichou County, Malipo County, Funing County, and Napo County (stream tributary to the upper reaches of the Shuikou River, same basin as the type locality) within China and northeastern Vietnam. Following the clarification of its taxonomic status, we have provided a redescription of B. zebroida.

The research on the Beaufortiapingi group is mostly in its initial stages; however, their situation is far from optimistic. These species have become popular ornamental fish in China. Aquarium trade operators reap substantial profits through the capture and sale of these fish from the wild, yet among them are those who act with disregard for sustainability; their harvesting practices are often destructive. Given their rheophilic nature, projects that obstruct rivers can easily lead to regional extinction. Their low pollution tolerance and sensitivity to changes in water quality also contribute to the significant reductions in population that many species are suffering. In the type locality of B. viridis sp. nov., Wuming District, Nanning City, most small tributaries have been modified into step-like reservoirs for water storage, and those near agricultural irrigation areas are polluted, rendering these areas unsuitable for their survival. Interestingly, a stable population was discovered inside a commercial eco-camping site. To satisfy consumers’ pursuit of “pristine nature,” contractors have left some river sections poor developed, providing a refuge for this species. We suggest that future efforts should focus on increasing attention to these species, conducting fundamental research, and further exploring their scientific and economic potential. Simultaneously, it is crucial to enhance habitat conservation awareness, scientifically plan, and develop sustainably, ensuring harmonious coexistence between humans and nature.

Beaufortia szechuanensis: SHOU20231220601, one specimen, 59.18 mm SL, from Wanzhou District, Chongqing, upper Yangtze River basin. SHOU20231220602, one specimen, 42.94 mm SL, from Qijiang District, Chongqing, upper Yangtze River basin. SHOU20231220640-642, three specimens, 46.17–54.93 mm SL, from Neijiang City, Sichuan, upper Yangtze River basin. Collected by Jing-Chen Chen and Hao-Tian Lei.

Beaufortia liui: SHOU20240225701, one specimen, 36.64 mm SL, from Huidong County, Sichuan, upper Yangtze River basin. Collected by Can Liu. IHASW83VI0424-425, two specimens, from Huidong County, Sichuan, upper Yangtze River basin, inspection only. IHW 2247, 2446, two specimens, syntype, inspection only.

Beaufortia niulanensis: SHOU20231220614-615, SHOU20231220617, three specimens, 41.38–43.66 mm SL, from type locality, Qujing City, Yunnan, upper Yangtze River basin. Yangtze River basin. Collected by Xin-Rui Pu, Lao Xing, and Lingzi.

Beaufortia polylepis: SHOU20231220643-644, two specimens, 40.06–41.02 mm SL, from downstream near the type locality, Kaiyuan City, Yunnan, upper Nanpan River (Pearl River system). Collected by Lin Yang. IHW 774330, one specimen, syntype, inspection only.

Beaufortia huangguoshuensis: IHASW83IV0049, one specimen, syntype, and fin rays were counted; the other data refers to Chen and Tang (2000).

Beaufortia intermedia: IHW 87IV516, one specimen, paratype, inspection only. Measurement data refers to Tang et al. (1997).

Beaufortia leveretti: SHOU20231212626-627, two specimens, 44.65–52.42 mm, from Fangchenggang City, Guangxi, collated by Qian-Yu Liang. SHOU20231212628-630, three specimens, 59.94–64.71 mm, from type locality, Baisha County, Hainan, collector unknown.

Beaufortia kweichowensis: SHOU20231212601-602, two specimens, 53.77–55.24 mm, from Congjiang County, Guizhou, Duliu River (Pearl River system). SHOU20231212614-616, three specimens, 44.38–47.77 mm, from Pingle County, Guangxi, Li River (Pearl River system). Collected by Jing-Chen Chen.

Beaufortia cyclica: IHASW75IV1417, one specimen, holotype, inspection only. Measurement data refers to the original description.

Beaufortia yunnanensis, Beaufortia fasciolata, Beaufortia triocellata, Beaufortia multiocellata, Beaufortia buas, Beaufortia daon, Beaufortia elongata, and Beaufortia loos: Referenced original descriptions (Li et al. 1998; Kottelat 2001; Nguyen and Nguyen 2005).