Specimens of the new species were captured in a fish survey conducted in June 2021. Among the 18 collected specimens, five were preserved in 95% ethyl alcohol for DNA extraction, and the remaining 13 specimens were fixed in 10% formalin for two days and then preserved in 70% ethyl alcohol for morphological examination. Two paratype specimens were deposited at the Zhejiang Museum of Natural History, Hangzhou City, Zhejiang Province, and the holotype and the remaining paratypes and alcohol-preserved specimens were deposited at Shanghai Ocean University, Shanghai City, China. Another eight congeneric species of mainland China, which were caught from other fish surveys, were also included for molecular analysis in this study. The three species (Formosania davidi, F. galericula and F. fasciolata) were used for morphological comparison with the new species because they are similar in morphology and geographically adjacent. The suffixes -Jiang and -Xi indicate rivers and streams, respectively, in Mandarin Chinese.

All measurements were taken point-to-point with a digital caliper and recorded to the nearest 0.1 mm, following Yi et al. (2014). Measurements and counts were made on the left side of the specimens whenever possible. Morphometric measurements were expressed as percentages of standard length (SL), head length (HL), or eye diameter. The numbers of fin rays and lateral-line scales were counted under a research microscope.

Genomic DNA was extracted from the alcohol-preserved pectoral-fin tip, and the mitochondrial cytochrome b (Cytb) gene was selected for amplification and sequencing. The Cytb gene was amplified by polymerase chain reaction in 25 μL reactions containing 9.5 μL of H₂O, 1 μL of each primer, 1 μL of template DNA, and 12.5 μL of Taq Master Mix (Sangon Co.,Ltd., Shanghai, China). Polymerase chain reaction (PCR) was performed at 95 °C predenaturing (3 min), then at 94 °C denaturing (30 s), 54 °C annealing (45 s), 72 °C extension (1 min) for 35 cycles, and 72 °C final extension (5 min). The primer pairs used for amplification and sequencing were L14724 (GACTTGAAAAACCACCGTTG) and H15915 (CTCCGATCTCCGGATTACAAGAC) (Xiao et al. 2001). Amplified products were subsequently purified and utilized for sequencing by a commercial sequencing company. The obtained sequences were spliced using Seqman from DNASTAR’s Lasergene (Burland 2000) and then checked by utilizing BLAST analysis in the GenBank database. After confirmation, the targeted sequences were submitted to the GenBank database (Table 1), and provided accession numbers.

We sequenced 28 Cytb gene sequences of Formosania and retrieved 8 Cytb gene sequences of Formosania from GenBank. Vanmanenia stenosoma and V. pingchowensis were selected as outgroups for molecular phylogeny analysis (Table 1). A multiple sequence alignment was prepared for all sequences using MEGA v.11.0 (Tamura et al. 2021). The genetic distances (p-distance with 1000 bootstraps) of the sequences among taxa were also calculated by using MEGA v.11.0. The best substitution models (TIM2+R3) for maximum likelihood (ML) and the best substitution model (GTR+G+I) for Bayesian inference (BI) were selected in ModelFinder (Kalyaanamoorthy et al. 2017) by Akaike’s information criterion (AIC). The phylogenetic trees were inferred using Bayesian inference (BI) and maximum likelihood (ML) approaches. Bayesian analyses were conducted using MrBayes (Ronquist et al. 2012). Four simultaneous Monte Carlo Markov chains were run for 2 million generations, with sampling one tree per 100 replicates for each run, and the first quarter of the trees were discarded as burn-in; the remaining trees from two independent runs were used to construct a consensus tree. The ML analyses were conducted using IQ-TREE (Nguyen et al. 2015) with a total of 20 000 bootstrap replications performed.

Family Gastromyzontidae Hora,1950

Genus Formosania Oshima,1919

 Formosania immaculata Sun, Zhou & Yang, sp. nov.

https://zoobank.org/E62FDC2E-148C-45B3-8BC1-2AD18C2B486A

Figs 1, 2A, 5

Type material

Holotype. SHOU2021060325, 87.9 mm total length (TL), 77.4 mm standard length (SL), adult collected by Jia-Jun Zhou and Wei Sun on June 28, 2021, in Wuyi County, Jinhua City, Zhejiang Province, China (28.7179°N, 119.4939°E; c. 882 m a.s.l.).

Paratypes. Twelve specimens from the same locality as the holotype, SHOU2021060326-060337, 43.2–68.7 mm SL, were collected by Wei Sun and Jia-Jun Zhou on June 28, 2021.

Description

Morphometric measurements for the specimens examined are given in Table 2. See Fig. 1A–C for lateral, dorsal, and ventral views of the body and Fig. 2A for its mouthpart structures.

Table 2.

Download as

CSV

XLSX

Morphometric measurements and meristic counts for Formosania immaculata sp. nov., F. davidi, F. fasciolata and F. galericula.

Characters	F. immaculata sp.nov. (N=13)			F. davidi (N=15)		F. fasciolata (N=8)		F. galericula (N=11)
	Holotype	Holotype+paratypes
		Range	Mean+SD	Range	Mean+SD	Range	Mean+SD	Range	Mean+SD
Standard length (mm)	77.4	43.2–77.4	61.7±7.80	58.5–75.4	67.8±6.08	50.4–73.4	58.9±7.85	50.5–67.3	59.0±4.63
% of standard length (SL)
Body depth	17.9	16.1–21.1	17.6±1.29	17.3–19.2	18.4±0.59	16.2–17.4	16.8±0.41	13.5–18.4	15.5±1.37
Head length	22.9	22.5–25.9	23.6±1.05	21.1–24.5	23.1±0.99	22.7–27.0	24.2±1.48	22.7–26.3	24.5±1.00
Head depth	12.6	11.5–13.9	12.7±0.72	11.5–13.4	12.5±0.55	12.5–13.4	13.0±0.29	11.8–13.3	12.5±0.48
Head width	18.7	17.4–19.7	18.3±0.61	16.0–19.5	17.8±1.07	18.1–19.8	19.0±0.61	13.5–20.0	18.0±1.99
Length of caudal peduncle	12.2	12.2–14.3	13.2±0.65	10.5–13.9	11.8±1.01	10.2–12.8	11.1±0.82	10.4–13.0	11.8±0.83
Depth of caudal peduncle	12.4	12.4–14.3	12.9±0.59	11.3–13.2	12.2±0.58	12.2–13.8	13.1±0.56	10.2–12.7	11.3±0.65
Dorsal-fin length	20.7	19.6–22.4	20.9±0.90	20.2–23.1	21.4±0.96	20.2–23.2	21.8±1.03	19.6–22.4	21.2±0.84
Pectoral-fin length	23.5	22.0–25.0	23.7±0.92	22.3–24.2	23.3±0.53	22.8–27.4	24.5±1.47	22.4–26.7	25.0±1.31
Pelvic-fin length	20.3	19.0–21.2	19.9±0.65	19.2–21.6	20.2±0.80	19.8–22.2	21.0±0.90	19.1–21.8	20.8±0.87
Anal-fin length	19.1	16.8–19.1	18.1±0.74	17.7–20.5	19.3±1.04	18.1–22.1	19.6±1.18	17.1–19.8	18.4±0.99
Dorsal-fin base length	12.6	12.3–13.8	12.8±0.46	11.3–14.1	12.8±0.85	12.6–15.3	13.7±0.91	10.5–13.6	12.4±0.83
Pectoral-fin base length	7.5	6.6–8.4	7.3±0.58	6.2–8.0	6.8±0.57	6.3–8.1	7.4±0.61	6.4–8.5	7.3±0.58
Pelvic-fin base length	5.1	4.6–5.4	5.1±0.26	4.3–5.3	4.7±0.32	4.4–5.7	5.2±0.44	4.6–5.5	5.2±0.29
Anal-fin base length	7.4	6.4–8.1	7.1±0.51	6.3–8.0	7.3±0.59	5.9–8.6	7.2±1.03	6.0–7.7	6.9±0.49
Predorsal length	47.5	45.4–50.5	48.2±1.45	47.3–49.9	49.0±0.96	46.7–51.2	48.8±1.37	47.6–49.8	49.1±0.72
Prepectoral length	18.2	18.1–20.9	19.5±0.96	17.2–20.8	19.3±1.00	20.2–22.2	21.0±0.69	17.8–23.1	20.7±1.36
Prepelvic length	52.1	50.8–54.5	52.8±1.07	51.7–55.0	53.4±0.94	52.2–54.9	53.8±0.95	52.0–55.8	54.2±1.42
Preanal length	79.3	78.1–80.4	79.4±0.74	77.2–82.8	79.9±1.91	78.4–83.3	80.6±1.71	75.9–82.2	79.8±1.94
% of head length (HL)
Snout length	47.8	44.6–48.2	46.0±0.93	40.7–44.8	43.2±1.35	42.2–43.8	43.2±0.49	39.9–45.9	43.1±1.96
Head depth	55.0	48.7–60.6	53.9±3.44	50.1–56.9	53.8±2.17	49.5–58.8	53.8±3.08	47.6–53.9	51.0±1.99
Eye diameter	13.9	13.7–16.8	15.1±1.07	13.1–15.8	14.3±0.90	14.8–17.7	16.3±0.92	15.1–17.3	15.9±0.81
Interorbital width	43.4	35.1–43.4	39.2±2.53	32.1–42.5	38.8±2.73	32.6–41.9	37.9±2.90	33.3–39.1	36.4±1.88
% of caudal peduncle length
Depth of caudal peduncle	101.7	91.0–101.9	97.8±3.74	86.0–109.3	103.4±7.34	107.9–130.6	118.8±7.85	86.2–106.2	96.5±6.01
% of eye diameter
the outermost pair of rostral barbels length	124.2	112.9–140.0	122.2±8.61	128.3–175.4	159.0±15.63	119.8–167.5	145.6±14.01	105.6–138.8	113.5±9.09
Maxillary barbels length	113.7	94.2–123.9	105.1±9.00	94.3–150.4	126.9±17.61	94.0–142.0	120.1±14.25	83.5–107.0	95.0±7.01
Lower lip papillae length	23.2	19.9–24.4	22.8±1.59	47.0–61.9	53.0±5.36	40.3–51.0	47.0±3.25	16.3–25.4	21.0±3.02
% of the postorbital length
Snout length	124.3	114.5–125.0	119.6±3.62	97.7–102.7	100.6±1.46	99.6–103.5	101.6±1.43	96.4–102.4	99.3±1.78
Meristic counts
Dorsal-fin rays	iii,8	iii,8		iii,8		iii,8		iii,8
Pectoral-fin rays	i,14	i,13–14		i,14–15		i,14		i,13–14
Pelvic-fin rays	i,8	i,8		i,8		i,8		i,8
Anal-fin rays	ii,5	ii,5		ii,5		ii,5		ii,5
Lateral-line scales	92	92–98		89–100		90–96		89–95

Figure 1. 

Lateral (A) dorsal (B) and ventral (C) views of Formosania immaculata sp. nov., holotype, adult, SHOU2021060325.

Figure 2. 

Ventral view of mouth of A Formosania immaculata sp. nov., SHOU2021060325, holotype B Formosania stigmata, SHOU2021060180. lj: lower jaw; llp: lower lip papilla; mb: maxillary barbel; rb: rostral barbel; rf: rostral fold; uj: upper jaw; ul: upper lip.

Head depressed in lateral view; head width always greater than depth; head width 17.4–19.7% of SL. Snout obtuse in dorsal view and longer than postorbital length; snout length 114.5%–125.0% of postorbital length. Mouth inferior and arched. Lips fleshy, with upper lip wide, without obvious convex particles; lower lip with a pair of papillae and a pair of lobulated papillae (Fig. 2A). Upper lip connected to lower lip around the corners of mouth by a papillated flap and one pair of maxillary barbels at the corners of mouth. Upper jaw covered by upper lip. Rostral fold appeared at end of snout, with 13 well-developed rostral barbels in an irregular row, covered with small spots. All rostral barbels connected with rostral fold. Outermost pair of rostral barbels longest and slightly longer than eye diameter, with a length of 112.9%–140.0% of eye diameter. Anterior and posterior nostrils adjacent with a well-developed flap on anterior ones. Eyes normal; diameter 13.7%–16.8% of HL. Gill openings reached ventral surface of head, with its upper extremity reaching the level of upper margin of orbit. Body elongated, the anterior part of body cylindrical and laterally compressed behind dorsal-fin base. The greatest depth of body at dorsal-fin origin and the least depth at caudal-fin base; body depth at dorsal-fin origin 16.1%–21.1% of SL. Body scaled but scales absent on head and before the middle of pectoral-fin base of abdomen. Scales minute, lateral line complete with 92–98 perforated scales. Caudal peduncle compressed laterally; length equal to peduncle depth.

Dorsal fin had three unbranched and eight branched rays; origin slightly in front of pelvic-fin insertion, situated slightly ahead to the midpoint between snout tip and caudal-fin base. Pectoral fin developed, with one unbranched and 13–14 branched rays. Pelvic fins long with one unbranched and eight branched rays, tips of depressed pelvic fins reaching anus when pelvic-fin rays extended backward. Anus in middle of pelvic-fin insertion and anal-fin insertion; anal fin with two unbranched and five branched rays, with the tip of anal fin closing or reaching to caudal-fin base. Caudal fin slightly forked; lower lobe slightly longer than upper lobe.

Coloration. In life, body slightly brown; fins and rostral barbels slightly red; with inconspicuous black blotches on back of head (Fig. 5). In 10% formalin-fixed specimens, dorsal and flank of head and body grayish-brown; ventral surface of head and abdomen white to yellowish with many black spots after pectoral fins; all fins hyaline and light gray, without obvious blotches (Fig. 1).

Diagnosis

Formosania immaculata sp. nov. resembles the Formosania davidi species group (F. davidi, F. galericula and F. fasciolata) in having 13 well-developed rostral barbels arranged in one irregular row (Fig. 2A), while other congeners arranged in 2 rows (Fig. 2B). It is distinguished from the three species in the Formosania davidi species group by having no obvious blotches or stripes (vs. having blotches or stripes) on the body and snout length longer than (vs. equal to) postorbital length (see Fig. 3 and Table 3). The new species differs from F. davidi in having shorter rostral barbels [outermost pair 112.9%–140.0% (average 122.2) vs. 128.3%–175.4% (average 159.0%) of eye diameter], shorter lower lip papillae [19.9%–24.4% (average 22.8) vs. 47.0%–61.9% (average 53.0%) of eye diameter] and narrower abdominal scaleless area (extending to middle of pectoral-fin base vs. extending slightly behind pectoral-fin axil) (see Table 3); from F. fasciolata in having shorter rostral barbels [outermost pair 112.9%–140.0% (average 122.2) vs. 119.8%–167.5% (average 145.6%)] of eye diameter, shorter lower lip papillae [19.9%–24.4% (average 22.8) vs. 40.3%–51.0% (average 47.0%) of eye diameter] and shorter depth of caudal peduncle [91.0%–101.9% (average 97.8%) vs. 107.9%–130.6% (average 118.8%) of caudal peduncle length]; and from F. galericula in having narrower abdominal scaleless area (extending to middle of pectoral-fin base vs. extending slightly behind pectoral-fin axil).

Table 3.

Download as

CSV

XLSX

Comparison of characters among Formosania immaculata sp. nov., F. davidi, F. fasciolata and F. galericula.

	F. immaculata sp. nov.	F. davidi	F. fasciolata	F. galericula
Distribution	Ou-jiang	Min-jiang	Feiyun-jiang	Ou-jiang
Blotches on the back	Absent	7–9 black blotches	7–9 light-colored blotches	7–13 black blotches
Blotches on the flank	Absent	Irregular blotches	18–22 anomalistic stripes	Filled with irregular blotches
Pectoral-fin rays	i,13–14	i,14–15	i,14	i,13–14
Outermost pair of rostral barbels length (% of eye diameter)	112.9–140.0 (122.2±8.61)	128.3–175.4 (159.0±15.63)	119.8–167.5 (145.6±14.01)	105.6–138.8 (113.5±9.09)
Lower lip papillae length (% of eye diameter)	19.9–24.4 (22.8±1.59)	47.0–61.9 (53.0±5.36)	40.3–51.0 (47.0±3.25)	16.3–25.4 (21.0±3.02)
Depth of caudal peduncle (% of caudal peduncle length)	91.0–101.9 (97.8±3.74)	86.0–109.3 (103.4±7.34)	107.9–130.6 (118.8±7.85)	86.2–106.2 (96.5±6.01)
Abdominal scaleless area	Extending to the middle of the pectoral-fin base	Extending slightly behind the pectoral-fin axil	Extending to the middle of the pectoral-fin base	Extending slightly behind the pectoral-fin axil
Snout length (% of the postorbital length)	114.5–125.0 (119.6±3.62)	97.7–102.7 (100.6±1.46)	99.6–103.5 (101.6±1.43)	96.4–102.4 (99.3±1.78)

Figure 3. 

Lateral and dorsal views of A Formosania davidi, SHOU2021060176 B Formosania fasciolata, SHOU2021060200 C Formosania galericula, SHOU2021060169.

Etymology

The specific epithet is the Latin form of the word immaculate here referring to the unique body of no blotches or stripes. We propose the Chinese common name Wú Bān Yīng Kǒu Qiū (无斑缨口鳅).

Distribution and habitat

The new species is known only from the upper reaches of the Xuanping-Xi, a stream tributary to the Ou-Jiang River, in Wuyi County, Zhejiang Province, China (Fig. 4). It inhabits fast-flowing streams with gravelly and pebbly substrates (Fig. 5).

Figure 4. 

Map showing collection localities of nine species of Formosania involved in the present study. The names of rivers are italicized, and the city of Jinhua is highlighted.

Figure 5. 

Habitat and live specimen of Formosania immaculata sp. nov.

Thirty-six Cytb gene sequences of Formosania were used for phylogenetic analysis. After alignment and trimming, 1141 bp (base pairs) of the Cytb gene was obtained. There were 854 conserved sites, 287 variable sites, 15 singleton sites, and 272 parsimony-informative sites. The mean frequency of four nucleotides was A=25.8%, T=28.2%, C= 30.0%, and G=16.0%; the base composition was A-T rich (54.0%).

The two phylogenetic analysis methods (BI and ML) showed an identical topology (Fig. 6). The monophyly of the genus was well supported (94% bootstrap value and 100% posterior probability). Both phylogenetic trees revealed that Formosania chenyiyui is sister to the remaining species of Formosania. The remaining nine species formed a monophyletic clade with strong support and then separated into three groups. Samples of F. immaculata sp. nov. were monophyletic and belonged to a group of species with robust support (100% bootstrap value and 100% posterior probability). This group is defined here as the F. davidi species group, and our molecular data suggest that it includes F. davidi, F. galericula and F. fasciolata.

Figure 6. 

Bayesian inference tree based on mitochondrial Cytb gene sequences of 10 Formosania species. Maximum likelihood and Bayesian inference analyses resulted in congruent trees. Bootstrap and posterior probability values are shown beside nodes on the tree if 50% or higher.

Formosania immaculata sp. nov. had minimal genetic distance with the three similar species, 4.5% with F. davidi, 4.5% with F. galericula, and 4.7% with F. fasciolata, which was greater than the genetic distance among the three similar species (2.5%–3.0%) (Table 4). The mean genetic distance of the new species from all sampled species was 8.6%, far greater than the minimum distance (2.3%), detected here between F. stigmata and F. paucisquama. Formosania chenyiyui has the greatest genetic divergence from all other species (16.2%–19.1%).

Formosania davidi: SHOU2021060096-106, SHOU2021060176-179, 15, 58.5–75.4 mm SL; Qingyuan County, Min-Jiang River System, Zhejiang Province, China.

Formosania fascicauda: SHOU202201083-091, 9, 53.4–69.1 mm SL; Nanjing County, Jiulong-Jiang River System, Fujian Province, China.

Formosania fasciolata: SHOU2021060193-200, 8, 50.8–73.4 mm SL; Liguang stream in Wuyanling National Nature Reserve (type locality), Taishun County, Feiyun-Jiang River System, Zhejiang Province, China.

Formosania galericula: SHOU2021060165-175, 11, 50.5–67.4 mm SL; unnamed stream in Hehu village (type locality), Qingyuan County, Ou-Jiang River System, Zhejiang Province, China.

Formosania paucisquama: SHOU202110011-013, 4, 50.3–64.4 mm SL; unnamed stream in Da’nan Mountain (type locality), Puning County, Lian-Jiang River System, Guangdong Province, China.

Formosania stigmata: SHOU2021060180-183, 4, 56.0–84.8 mm SL; Qingyuan County, Min-Jiang River System, Zhejiang Province, China.

Formosania tinkhami: SHOU202110086-090, 5, 44.6–58.1 mm SL; unnamed stream in Nankun Mountain (type locality), Longmen County, Pearl River System, Guangdong Province, China.

The authors have declared that no competing interests exist.

No ethical statement was reported.

This research was funded by a grant from the National Natural Science Foundation of China (No. 31872207).

Sun W contributed to the experimental work, data analysis and drafting of the manuscript. Zhou JJ performed specimen collection and photograph. Yang JQ was responsible for review and editing.

Jia-Jun Zhou https://orcid.org/0000-0003-1038-1540

Jin-Quan Yang https://orcid.org/0000-0003-0387-1824

All of the data that support the findings of this study are available in the main text.