Laubuka tenella, a new species of cyprinid fish from southeastern Bangladesh and southwestern Myanmar (Teleostei, Cyprinidae, Danioninae)

Abstract

Laubuka tenella is a new species characterized by the colour pattern, consisting of short dark vertical bars anteriorly on the side, and a dark lateral band posteriorly on the side, combined with a relatively short pelvic fin and 29–30 lateral-line scales. It is separated from other Laubuka analysed by minimum 9 % uncorrected p -distance in the mitochondrial COI gene. The type series is composed of specimens from small streams in the Cox’s Bazar District in Bangladesh (the type locality), and the Thandwe River drainage in western Myanmar. Laubuka brahmaputraensis is strongly indicated to be a junior synonym of L. laubuca , the second known species of Laubuka in Bangladesh. Eustira ceylonensis , currently in the synonymy of Devario malabaricus , is a valid species of Laubuka .

Full text

Laubuka tenella, a new species of cyprinid sh from southeastern Bangladesh... 105
Laubuka tenella, a new species of cyprinid fish from
southeastern Bangladesh and southwestern Myanmar
(Teleostei, Cyprinidae, Danioninae)
Sven O. Kullander1, Md. Mizanur Rahman2, Michael Norén1, Abdur Rob Mollah2
1 Department of Zoology, Swedish Museum of Natural History, PO Box 50007, SE-104 05 Stockholm, Sweden
2 Department of Zoology, University of Dhaka, Dhaka-1000, Bangladesh
Corresponding author: Sven O. Kullander (sven.kullander@nrm.se)
Academic editor: N.Bogutskaya|Received 24 November 2017|Accepted 8 February 2018|Published 12 March 2018
http://zoobank.org/853119CC-7CBD-4613-A120-50F9A328BEEA
Citation: Kullander SO, Rahman MM, Norén M, Mollah AR (2018) Laubuka tenella, a new species of cyprinid
sh from southeastern Bangladesh and southwestern Myanmar (Teleostei, Cyprinidae, Danioninae). ZooKeys 742:
105–126. https://doi.org/10.3897/zookeys.742.22510
Abstract
Laubuka tenella is a new species characterized by the colour pattern, consisting of short dark vertical
bars anteriorly on the side, and a dark lateral band posteriorly on the side, combined with a relatively
short pelvic n and 29–30 lateral-line scales. It is separated from other Laubuka analysed by minimum
9 % uncorrected p-distance in the mitochondrial COI gene. e type series is composed of specimens
from small streams in the Cox’s Bazar District in Bangladesh (the type locality), and the andwe River
drainage in western Myanmar. Laubuka brahmaputraensis is strongly indicated to be a junior synonym
of L. laubuca, the second known species of Laubuka in Bangladesh. Eustira ceylonensis, currently in the
synonymy of Devario malabaricus, is a valid species of Laubuka.
Keywords
DNA barcode, morphometry, phylogeny, South Asia, taxonomy, vertebral counts
Introduction
Species of the tropical Asian cyprinid genus Laubuka Bleeker, 1860, are characterized
by relatively small size, up to about 60 mm SL, a strongly compressed body and a
keeled abdomen, a long and sti falciform pectoral n, and usually a long free distal
ZooKeys 742: 105–126 (2018)
doi: 10.3897/zookeys.742.22510
http://zookeys.pensoft.net
Copyright Sven O. Kullander et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC
BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
RESEARCH ARTICLE
Launched to accelerate biodiversity research
A peer-reviewed open-access journal

Sven O. Kullander et al. / ZooKeys 742: 105–126 (2018)
106
portion of the rst pelvic-n ray (Pethiyagoda et al. 2008, Silas 1958). Laubuka re-
sembles Devario Heckel, 1843, in presence of supraorbital and rostral neuromast-lled
grooves, posteriorly situated dorsal n, and presence of a dark blotch immediately
posterior to the gill opening. Unlike Devario, most species of Laubuka show only in-
distinct markings, and barbels have not been recorded from Laubuka.
Species of Laubuka have been reported from southern Asia in Bangladesh, Cam-
bodia, India, Indonesia, Laos, Malaysia, Myanmar, Pakistan, Sri Lanka, Vietnam,
and ailand (Silas 1958, Kottelat 2013). About 13 species are recognized at present
(Pethiyagoda et al. 2008, Knight 2015, Lalramliana et al. 2017), but the genus has
not been subject to revision since that of Silas (1958), who employed a very dier-
ent concept of taxa currently in Laubuka. Banarescu (1968) examined and illustrated
several species of Laubuka, but his paper is mainly concerned with the description of a
new subgenus, Malayochela, monotypic with M. maassii (Weber & de Beaufort, 1912).
Banarescu’s concept of the other species of Laubuka is similar to that of Silas (1958).
Several of the species presently in Laubuka were previously, e.g., in Silas (1958) and
Banarescu (1968), in the genus Chela (Hamilton, 1822), from which they were sepa-
rated by Pethiyagoda et al. (2008). Species diversity in Chela was addressed by Knight
and Rema Devi (2014), who recognized two valid species in that genus. Another genus
previously included as a subgenus of Chela is Neochela Silas, 1958, containing only the
miniature species N. dadiburjori (Menon, 1952). Malayochela was considered to be a
distinct genus by Pethiyagoda et al. (2008).
e type species of Laubuka is Cyprinus laubuca Hamilton, 1822, which was de-
scribed from “ponds of the northern parts of Bengal” (Hamilton 1822), i.e., within
the present state of West Bengal in India, and Bangladesh. Hamilton (in Day, 1877)
reporting on surveys 1807–1813, specically mentioned the species from Goalpara in
Rangpur (currently Goalpara District, Assam, India, Brahmaputra basin), Mahananda
River in Purniah (Purnea District, Bihar, Ganga basin), and in Gorakhpur District
(Uttar Pradesh, Ganga basin), suggesting that he did not observe the species in locali-
ties within present-day Bangladesh. Hamilton left India in 1815 (Day, 1877), however,
and would have had opportunities to make unrecorded observations of shes. Silas
(1958), followed by Talwar and Jhingran (1991), reported L. laubuca as distributed
in Pakistan, India, Bangladesh, Sri Lanka, Myanmar, Malay Peninsula, and Sumatra.
Recent publications only list specimens from India (Knight 2015, Pethiyagoda et al.
2008) and Bangladesh (Rahman and Chowdhuri 2007), but there is no recent revi-
sion covering this species. Silas’s (1958) concept of L. laubuca included the southeast
Asian L. siamensis Fowler, 1939, and the Sri Lankan L. lankensis (Deraniyagala, 1960),
revalidated by Pethiyagoda et al. (2008), and would have accommodated also the three
Sri Lankan species described by Pethiyagoda et al. (2008) as L. insularis, L. ruhuna,
and L. varuna.
Despite previous records from Bangladesh, we were unable to nd Laubuka in nat-
ural habitats in the Meghna, eastern Brahmaputra, and Karnafuli drainages in Bangla-
desh During eld work 2014– 2016. Laubuka laubuca was, however, present in small
numbers in aquarium shops in Dhaka, and shop owners said that they had been caught

Laubuka tenella, a new species of cyprinid sh from southeastern Bangladesh... 107
locally. In 2015, we collected specimens of Laubuka from streams near Cox’s Bazar
and Teknaf in extreme southeastern Bangladesh. ese samples represent a species very
dierent from L. laubuca, but morphologically indistinguishable from a sample of
Laubuka from a stream on the western slope of the Rakhine Yoma in Myanmar. is
paper is dedicated to the description and diagnosis of this new species.
Materials and methods
Specimens used were available in museums, purchased from shermen or markets; or
collected in the wild using beach seine or hand net and euthanized through immersion
in buered tricaine-methanesulphonate (MS 222) until cessation of opercular move-
ments plus an additional 30 minutes, in accordance with permits from the Swedish
Environmental Protection Agency (dnr 412-7233-08 Nv) and the Stockholm Ethical
Committee of the Swedish Board of Agriculture (dnr N 85/15). Collecting in Bang-
ladesh was conducted under a permit to the University of Dhaka. Voucher specimens
are deposited in the collections of e Natural History Museum, London (BMNH),
University of Dhaka, Dhaka (DU), Museum of Zoology, Lund University (MZLU),
and the Swedish Museum of Natural History, Stockholm (NRM).
Measurements and counts were taken as described by Fang (1997) with the ex-
ception that body depth was taken at level of the origin of the anal-n, which is very
slightly anterior to the origin of the dorsal n. Counts of lateral-line scales do not
include perforated scales on the caudal-n base. e last two dorsal- and anal-n rays
share the same proximal pterygiophore and may appear as a single ray; these two rays
are counted as 1½. Vertebral counts are given as precaudal+caudal, where the rst
vertebra bearing a haemal spine (anterior to the rst long anal-n pterygiophore) was
recorded as the rst caudal vertebra. Digital X-radiographs made with a Kevex 130kVP
microfocus X-ray source and a Samsung/Rayence 17×17 inch DR panel. Statistics were
calculated using SYSTAT v. 13 (Systat Software 2009).
For the genetic analysis, 655 basepairs from the 5’ end of the mitochondrial cy-
tochrome-oxidase subunit 1 (COI, or COX1) gene were sequenced from seven mor-
phologically identied specimens of Laubuka plus two specimens of the closely re-
lated Chela cachius (Hamilton, 1822). DNA was extracted using a ermo Scientic
KingFisher Duo (ermo Fisher Scientic, Waltham, USA) fully automated liquid-
handling instrument, with the ermo Scientic KingFisher Cell and Tissue DNA
Kit (ermo Fisher Scientic, Waltham, USA) with the recommended protocol. PCR
were performed with the puReTaq Ready-To-Go PCR kit (Amersham Biosciences AB,
Uppsala, Sweden). e COI fragment was amplied using the sh barcoding prim-
ers Fish-F1 and Fish-R1 [26], with the PCR cycling: 94 °C 4 min; 35 * (94 °C 30 s;
52°C 30 s; 72 °C 30 s); 72 °C 8 min). e PCR products were checked on agarose
gel, and puried by adding 5 µL of a mix consisting of 20 % Exonuclease I (EXO)
and 80% FastAP ermosensitive Alkaline Phosphatase (Fermentas/ermo Fischer
Scientic, Gothenburg, Sweden) to each 25 µl PCR reaction, incubated at 37°C for

Sven O. Kullander et al. / ZooKeys 742: 105–126 (2018)
108
30minutes, then heated to 80 °C for 15 minutes. Sequencing of both strands of all
fragments was carried out by Macrogen Europe (Amstelveen, e Netherlands). Se-
quences were assembled, aligned, and proofread in Geneious version 10 (Kearse et
al. 2012). Geneious was used to calculate genetic distances (uncorrected pairwise p-
distance, as recommended by Srivathsan and Meier (2011)), and the Geneious plug-in
Species Delimitation (Masters et al. 2011) was used to calculate the probability of re-
ciprocal monophyly under a model of random coalescence. A phylogenetic hypothesis
was constructed in MrBayes version 3.2 (Huelsenbeck and Ronquist 2001; Ronquist
and Huelsenbeck 2003) (2 million generations, GTR + Γ + I model), data partitioned
by codon position; samples were taken every 1000 generations, and the rst 25 % of
samples were discarded as ‘burn-in’. Convergence was checked with Tracer version 1.6
(Rambaut et al. 2014).
e distribution map was constructed using layers from Natural Earth (http://
www.naturalearthdata.com).
Comparative material. Chela cachius: DU 6116, 1, 40.9 mm SL. Bangladesh:
Chittagong Division: Feni River drainage, Kohua River; M.M. Rahman, 29 May
2015. — NRM 40494. 4, 46.8–49.3 mm SL; India: Assam: about 100 km SE of Di-
brugarh, small river falling into the Dibru River 3 km north of Digholtarang; F. Fang
& A. Roos, 22 Jan 1998. — NRM 66988 (T10085), 1, 35.7 mm SL. Bangladesh:
Dhaka Division: Padma River near Sreenagar, M.M. Rahman et al., 2 Dec 2014.
Devario cf. malabaricus. BMNH 1852.2.19.130–132, 1853.12.24.6, 1864.4–
11.33; 6, 50.1– 67.2 mm SL; Ceylon.
Laubuka laubuca: Bangladesh: NRM 67315 (T10453), 1, 35.8 mm SL; NRM
67316 (T10454), 1, 31.9 mm SL; NRM 67317 (T10455), 1, 42.7 mm SL. Bang-
ladesh, ornamental sh shops in Dhaka, said to be local sh; M.M. Rahman et al.,
16May 2015. India, Assam: NRM 40308, 1, 48.7 mm SL; NRM 40486, 1, 44.0
mm SL; NRM 40517, 55.6–58.3, Assam, about 100 km SE of Dibrugarh, rivulet
falling into the Dibru River 3km N of Digholtarang, F. Fang & A. Roos, 22 Jan
1998. – NRM 44997 (T10965), 38.9 mm SL; Brahmaputra River at Tezpur, Sonit-
pur; H.Bleher, 28 Feb 2009. — NRM 52691, 1, 43.8 mm SL; Golaghat: Kaziranga,
Deosor subdrainage; O. Åhlander et al., 24 Oct 2005. — NRM 57234, 2, 49.9–53.1
mm SL; Dibrugarh, Mr Kamal Lahkar’s home sh farm “Brahmaputra Aquaria”.
F.Fang& A.Roos, 19 Jan 1998.
Laubuka sp.: NRM 12218, 59.0 mm SL; India: Kerala: Meenachil River and
adjacent canals, NW of Kottayam; E. Åhlander et al., 5–6 Dec 1987.
Laubuka parafasciata Lalramliana, Valalhlimpuia & Singh, 2017: All from
Myanmar, Rakhine State. BMNH 2017.8.2.2, 1, 45.6 mm SL; Kananme Chaung near
Leldee village and Chaung Ma Gyi Chaung, Leldee village, 18°35.814'N, 94°22.853'E
and 18°35.112'N, 94°22.182'E; R. Britz, 29 Nov 2009. — BMNH 2017.8.2.3-
4, 2, 64.7–85.3 mm SL; Chaung Ma Gyi Chaung, Leldee village, N 18°35.112'N,
94°22.182’E; R. Britz, 28 Nov 2009. — BMNH 2017.8.2.21, 1, 37.8 mm SL; Kanan-
mee Chaung; Ye Hein Htet, 3 Dec 2004. — BMNH 2017.8.2.23, 1, 46.9 mm SL;
Ann Chaung near Ann; R. Britz, 24–25 Nov 2009. (Ann = 19°47’0"N, 94° 2'0"E).

Laubuka tenella, a new species of cyprinid sh from southeastern Bangladesh... 109
Laubuka insularis: All from Sri Lanka. MZLU 962/5429, 2, 20.3–33.3 mm SL;
Eastern: Rambukkan Oya River, 25 miles NE of Bibile. P. Brinck et al., 8 Mar 1962.
(125). — MZLU 962/5442. 6, 32.9–44.0 mm SL; Eastern: Gal Oya River, 14 miles
E of Bibile; P. Brinck et al., 8 Mar 1962. (123). — MZLU 962/5459, 1, 47.8 mm SL;
North Central: Mahaweli River drainage: small stream 3 mi. S of Minneriya; P. Brinck
et al., 11 Feb 1962 (67).
Laubuka cf. varuna: MZLU 962/5438, 4, 29.3–46.0 mm SL; Sri Lanka: Central:
Talagala Oya stream at Pidurutalagala, 2 miles N of Nuwara-Eliya; P. Brinck et al., 4
Mar 1962. (116:I).
Laubuka siamensis: NRM 31223, 6. Viet Nam; Dong Nai River drainage,
about 30 km ENE of Ho Chi MinhCity, Xa Trang Bom, road crossing small stream
east of village, 10°57'1"N, 106°58'28"E; B. Björkegren, 25 Mar 1935. Information
on L. ruhuna, L. varuna, L. lankensis from Pethiyagoda et al. (2008); on L. khujairo-
kensis (Arunkumar, 2000) from Arunkumar (2000); on, L. caeruleostigmata (Smith,
1931) from Smith (1931); and on L. trevori Knight, 2015 and L. latens (Knight,
2015 from Knight (2015).
GenBank Accession numbers. New COI sequences generated for this paper are:
Chela cachius, DU 6116: MG895632; NRM 66988: MG895633.
Laubuka laubuca, NRM 67315: MG895634; NRM 67317: MG895635; NRM
44997: MG895636.
Laubuka tenella: NRM 67380: MG895640; DU 9008/NRM 67381: MG895637;
NRM 67862: MG895639; DU 9006/NRM 67845: MG895638.
Results
Laubuka tenella sp. n.
http://zoobank.org/93EB4FA4-7823-4951-A3A7-A0F99BAF3672
Figs 1–3
Holotype. (Fig. 1A). DU 9004, 42.1 mm SL. Bangladesh: Chittagong Division: Cox’s
Bazar District: Naf River drainage, Domdomia stream, 10 km north of Teknaf town, 70
km south of Cox’s Bazar; 20°55'24"N, 92°15'47"E; M.M. Rahman et al., 9 May 2015.
Paratypes. DU 9005, 6, 30.8–42.2 mm SL; DU 9006/NRM 67845, 1, 47.4 mm
SL; DU 9007/NRM 67861, 1, 35.8 mm SL; NRM 67862, 1, 38.4 mm SL; NRM
69227, 1, 47.4 mm SL; NRM 68062, 7, 33.9–45.7 mm SL; same data as holotype.
NRM 67380, 1, 46.8 mm SL; DU 9008/NRM 67381 , 1, not measured; Bangladesh,
Chittagong Division: Bakkhali River drainage, Majerchora stream, 10 km south of
Cox’s Bazar; 21°23'45"N, 92°0'16"E; M.M. Rahman et al., 8 May 2015. BMNH
2018.1.31.3–5, 3, 37.9 mm SL; NRM 40813, 15, 33.9–42.5 mm SL; NRM 69798,
10, 29.6–37.0 mm SL; Myanmar: Rakhine State; andwe River drainage: Nan
Chaung, a stream at 3 km on road from andwe (market) to Ngapali; 18°27'8"N,
94°20'55"E; S.O. Kullander & R. Britz, 20 Mar 1998.

Sven O. Kullander et al. / ZooKeys 742: 105–126 (2018)
110
Figure 1. Laubuka tenella. A holotype, DU 9004, 42.1 mm SL. Bangladesh: Chittagong Division:
Domdomia stream, 10 km north of Teknaf B Laubuka tenella, paratype, NRM 40813, 39.8 mm SL.
Myanmar: Rakhine State: andwe River drainage, Nan Chaung, near andwe C Laubuka tenella,
paratype, NRM 67380, 46.8 mm SL, preserved in 95% ethanol. Bangladesh: Chittagong Division,
Majerchora stream, 10 km south of Cox’s Bazar.
Diagnosis. Distinguished from all other species of Laubuka except L. insularis,
L.lankensis, L. ruhuna, and L. varuna by the colour pattern, including a dark stripe
along the middle of the posterior third of the side or slightly shorter, anteriorly replaced
by 6–11 short vertical bars (vs. presence of a dark stripe along the side but absence of
bars in L. fasciata, L. parafasciata, and L. trevori; plain sides or presence of a very nar-
row posterior stripe in L. khujairokensis, L. latens, and L. laubuca; indistinct vertical
bars anteriorly on the side, followed by a dark stripe ending in a triangular spot on the
caudal-n base in L. siamensis; a few dark bars present anteriorly on the side but lateral
band absent in L. caeruleostigmata). Distinguished from L. fasciata, L. latens and L. tre-
vori also by more dorsal-n rays (ii.8½ vs. ii.7½) and from L. siamensis by the absence
of a dark spot on the caudal-n base. Distinguished from L. insularis by fewer scales in
the lateral line (29 –32 vs. 34 –36), shorter pelvic n (not reaching to vent, vs. reach-
ing to bases of anal-n rays 3–8); from L. lankensis by fewer scales in the lateral line
(29–32vs. 34–37); from L. ruhuna and L. varuna by the presence of an entire, narrow
lateral band on the posterior third of the body, vs. a series of blotches along the side
which may form a broad band extending anteriorly to about the middle of the side.

Laubuka tenella, a new species of cyprinid sh from southeastern Bangladesh... 111
Figure 2. Laubuka tenella, paratype from the type locality, photographed alive immediately upon cap-
ture; specimen preserved, but not possible to match with particular preserved specimen.
Description. Elongate, strongly compressed laterally. Predorsal contour slightly
ascending, levelling out close to dorsal n-base, minor indentation at commencement
of squamation. Dorsal-n origin marking 2/3 of standard length, immediately poste-
rior to vertical from anal-n origin. Dorsal-n base contour slanting, continuous with
dorsal contour of caudal peduncle; caudal peduncle only slightly tapering caudad.
Snout shorter than orbital diameter, triangular in lateral aspect, rounded in dor-
sal aspect. Mouth terminal, lower jaw at about 50° angle, tip anterior to upper jaw,
not quite reaching level of upper margin of orbit. Eyes large, lateral, in middle of
head length, well visible in ventral aspect of head, in dorsal aspect only slightly.
Anterior nostril tubular, opening anterolaterad. Supraorbital ending anteriorly in
sharp point. Long shallow frontal and rostral neuromast grooves present. Barbels
absent. Tubercles absent. Lower jaw with wide band of minute papillae, tentatively
identied as neuromasts. Ventral outline more arched than dorsal; slanting about
straight to under pectoral-n base, posteriorly about straight horizontal to anal-n
insertion; anal-n base contour straight ascending. Chest at close to isthmus; from
pectoral-n base caudad to pelvic-n base strongly compressed, posteriorly strongly
compressed and with sharp keel formed by margins of opposed left and right side
abdominal scales.
All scales cycloid, thin, transparent. Lateral line anteriorly descending for about ve
scales, posteriorly paralleling ventral outline, ending on lower half of caudal peduncle,
continued by 1–2 scales basally on caudal n. Single row of scales along base of anal
n. Elongate axillary pelvic-n scale present. Lateral line scales 29 (1), 30 (2), 31 (7),
32 (3) in Cox’s Bazar specimens; 29 (1), 30 (5), 31 (7) in Rakhine specimens. Predorsal
scales 16 (1), 17 (6), 18 (4) in Cox’s Bazar specimens; 16 (2), 17 (7), 18 (1) in Rakhine
specimens. Circumpeduncular scales 12 (27). Scales between dorsal n origin and lat-
eral line ½6 (28); between lateral line and anal-n origin 3 (28), of which distal scale
part of abdominal keel.
Dorsal-n origin at about posterior third of body, slightly posterior to origin of
anal n; with straight distal margin, rays gradually shorter from second unbranched
ray to last ray. Dorsal-n rays ii.8½ (29). Anal n with longer base than dorsal n;
short rounded anterior lobe, posterior rays gradually shorter. Anal-n rays iii.16½ (3),
iii.17½ (3), iii.18½ (8) in Cox’ Bazar specimens; iii.18½ (3), iii.19½ (10, iii.20½ (2) in

Sven O. Kullander et al. / ZooKeys 742: 105–126 (2018)
112
Figure 3. Body depth plotted against standard length in Laubuka laubuca and L. tenella.
Rakhine specimens. Pectoral-n long, falcate, unbranched ray longest or unbranched
and rst branched ray equally long, not reaching to vent; two large scales covering
bases of branched n rays and adjacent chest. Caudal n forked to about middle of n.
Pectoral-n rays i.10 (1), i.11 (10), i.11 (3) in Cox’s Bazar specimens; i.10 (4), i.11 (9),
i.12 (2) in Rakhine specimens. Pelvic n inserted slightly anterior to middle of side;
short, unbranched ray longest with short prolongation, not reaching to vent. Pelvic-n
rays i.5 (1), i.6 (13) in Cox’s Bazar specimens; i.6 (15) in Rakhine specimens.
Vertebrae: predorsal 16 (2), 17 (5), precaudal+caudal 15+18 (1), 15+19 (5), 16+19
(1), within caudal peduncle 5 (4), 6 (3) in Cox’s Bazar specimens; predorsal 16 (1), 17
(5), precaudal+ caudal 15+18 (1), 15+19 (4), 16+19 (1), within caudal peduncle 5 (4),
6 (2) in Rakhine specimens.
Ground colour in formalin-xed specimens (Figs 1A–B) pale yellowish white with
diuse grey or black markings except for round black cleithral spot size of pupil. Dor-
sum sparsely pigmented; brown stripe on dorsal midline from occiput to end of caudal
peduncle. in black or brown stripe along middle of caudal peduncle or slightly long-
er, anteriorly replaced by 6–11 grey or brown short vertical bars, not reaching ventrally
onto abdomen, less distinct in some larger specimens. A few black dots along middle
of abdomen. Fins hyaline. Ethanol-xed specimens (Fig. 1C), with silvery opercle and
sides; vertical bars indistinct, dorsum grey, abdominal sides pale yellow; cleithral spot
and lateral band black. Live specimens observed only in the type locality , with silvery
reections dorsally, abdomen white, sides blue, along middle a wide iridescent green
cleithral spot anked by golden (Fig. 2).

Laubuka tenella, a new species of cyprinid sh from southeastern Bangladesh... 113
Etymology. e specic name is a Latin adjective in diminutive form, tenellus,
here in the meaning of delicate, referring to the small size and the soft, delicate consist-
ency of fresh specimens.
Comparative morphometry. Laubuka tenella is slightly more slender than L.laubuca
of similar size (Tables 1–2, Fig. 3), but size dierences and potential sexual dimorphism
between the measurement series prevent a conclusive comparison. e sample of Laubuka
laubuca (N=7) is too small to establish signicant linear regression parameters.
Phylogenetic characterization and relationships. e Bayesian phylogenetic
analysis recovered Laubuka tenella as the sister species of L. laubuca (Fig. 4). COI
sequences of L. tenella diered from the most similar sequences, in L. laubuca, by
9 % uncorrected p-distance. e within-species variation in L. tenella amounted to
0.9 %, between the Majerchora and Domdomia samples. Maximum pairwise p-dis-
tance in L. laubuca was 2.5% when the sequence KT353103 was included, and 0.9 %
when excluded. Species delimitation methods conrmed reciprocal distinctness of L.
laubuca and L. tenella: P ID(Liberal), the mean probability of making a correct identi-
cation of an unknown specimen of the focal species was reciprocally 0.97. At 6*10-4,
Rosenberg’s P(AB) failed the null hypothesis that the combined clade (L. laubuca+L.
tenella) represents a single species.
Geographical distribution and habitat. Laubuka tenella is known only from
small streams in the vicinity of Cox’s Bazar and Teknaf in Bangladesh, and andwe
in Myanmar (Fig. 5).
Table 1. Morphometry of Laubuka tenella. Measurements are in per cent of standard length, except for
standard length (in mm). SD, standard deviation; r, Pearson’s correlation coecient; linear regression
parameters calculated from measurements in mm, when ANOVA o= 0, and r> 0.9. HT = Holotype.
Measurements N HT Min Max Mean SD r(SL) slope (b) intercept (a)
SL (mm) 29 42.1 30.8 45.7 37.8 3.8 – – –
Body depth 28 25.2 22.4 28.0 25.3 1.4 0.93 0.306 -1.989
Head length 28 20.9 20.9 23.9 22.9 0.7 0.94 0.194 1.335
Snout length 28 6.4 6.0 7.9 7.0 0.5 0.76 – –
Head depth 28 14.7 13.6 15.8 14.7 0.6 0.91 0.128 0.698
Head width 28 11.9 11.8 13.3 12.6 0.4 0.95 0.100 0.968
Upper jaw length 28 7.1 7.1 8.6 7.8 0.4 0.86 0.061 0.607
Lower jaw length 28 10.7 9.8 13.0 11.2 0.8 0.74 0.069 1.616
Orbit diameter 28 7.6 7.2 9.1 8.3 0.5 0.81 – –
Interorbital width 28 10.9 10.2 12.0 11.3 0.5 0.91 0.079 1.264
Caudal-peduncle length 28 15.4 12.6 16.2 14.1 0.9 0.80 – –
Caudal-peduncle depth 28 9.0 8.2 10.8 9.3 0.6 0.89 0.110 -0.625
Dorsal-n base length 28 14.3 11.3 14.4 12.7 0.8 0.83 – –
Anal-n base length 28 23.8 23.0 26.4 24.7 1.0 0.93 0.266 -0.721
Predorsal length 28 65.6 64.7 72.7 67.8 1.8 0.97 0.694 -0.619
Preanal length 28 61.8 61.7 69.1 64.6 1.6 0.98 0.701 -2.066
Prepelvic length 28 41.6 41.6 45.5 43.9 1.0 0.97 0.397 1.586
Pectoral-n length 28 34.0 32.9 38.9 36.3 1.7 0.86 0.265 3.688
Pelvic n length 28 16.4 14.9 19.1 17.1 0.9 0.87 0.159 0.439

Sven O. Kullander et al. / ZooKeys 742: 105–126 (2018)
114
Figure 4. Phylogram of relationships of Laubuka tenella and similar taxa, based on a Bayesian analysis
of the mitochondrial COI gene. Branch lengths are proportional to expected changes per site, visualizing
estimated genetic distance. e scale bar represents number of expected substitutions per nucleotide site.
Node labels show the Bayesian posterior probability of the clade. Terminal labels start with GenBank ac-
cession number and end with a locality indication when known. Devario xyrops and Malayochela maassi
are outgroup taxa. HM224171, identied as Laubuka fasciata in GenBank, is apparently a misidentied
L.laubuca. JN815300 and JN815301 with locality in the Bay of Bengal o Bangladesh and India, obvi-
ously in error. CTOL samples lack locality information.
Collections were made in the dry season when the streams had very little water.
e type locality (Fig. 6) was in the lower course of the Domdomia stream, close to the
mouth of the Naf River, owing out of low forest into pasture associated with a village.
At the time the stream was very shallow, not more than 10 m wide, with slightly turbid
water and a bottom substrate of clay mixed with stones. e steep banks suggested that
the water level in the monsoon season would reach a few meters higher. e associated
sh fauna included Aplocheilus panchax (Hamilton) (Aplocheilidae), Megalops cyprinoides
(Broussonet) (Megalopidae), Acentrogobius caninus (Valenciennes) (Gobiidae), Eleotris
melanosoma Bleeker (Eleotrididae), Dermogenys burmanica Mukerji (Zenarchopterdae),

Laubuka tenella, a new species of cyprinid sh from southeastern Bangladesh... 115
Table 2. Morphometry of Laubuka laubuca. Measurements are in per cent of standard length, except for
standard length (in mm). SD, standard deviation; r, Pearson’s correlation coecient.
NMin Max Mean SD r(SL)
8 38.9 55.6 48.6 5.9
7 27.7 33.9 31.2 1.9 0.83
7 22.1 23.5 23.0 0.6 0.97
7 6.7 7.3 6.9 0.2 0.92
7 13.4 16.4 15.1 1.0 0.94
7 12.2 13.7 12.7 0.5 0.97
7 6.8 8.0 7.5 0.4 0.92
7 9.6 11.8 10.8 0.8 0.91
7 7.5 8.7 8.3 0.4 0.90
7 11.1 12.7 11.7 0.6 0.94
7 11.7 13.4 12.3 0.6 0.88
7 9.2 10.9 10.0 0.7 0.96
7 11.7 14.2 13.5 0.9 0.93
7 23.4 28.3 26.3 1.5 0.95
7 65.9 68.7 67.7 1.0 0.99
7 63.5 69.2 67.0 2.0 0.98
7 42.5 47.8 44.5 2.0 0.97
7 33.9 38.8 36.5 1.9 0.96
7 15.7 26.9 22.7 3.8 0.69
Figure 5. Map of collecting sites of Laubuka laubuca and L. tenella.

Sven O. Kullander et al. / ZooKeys 742: 105–126 (2018)
116
and Oryzias cf. dancena (Hamilton) (Adrianichthyidae), reecting proximity to the Naf
River estuary. Other associated species in the Domdomia stream were identied as Glos-
sogobius giuris (Hamilton) (Gobiidae), Channa gachua (Hamilton), C. punctata (Bloch)
(Channidae), Danio sp., Devario anomalus Conway, Mayden & Tang, Devario coxi Kul-
lander, Rahman, Norén & Mollah, Esomus danrica (Hamilton), Pethia ticto (Hamilton),
Puntius chola (Hamilton), and Rasbora rasbora (Hamilton (Cyprinidae).
e Majerchora stream, in a hilly landscape with low forest near the sea, was very
small, less than 1 m wide, and not more than about 30 cm deep. e water was only
slightly turbid, owing slowly over a bottom of sand and clay. Very few sh specimens
were collected at this site; associated species were identied as Danio sp., Devario coxi,
and Pethia ticto (Cyprinidae).
e locality in Myanmar was a stagnant pool in a desiccated small river, with leaf
litter and sand on the bottom. e associated species were identied as Anguilla sp.
(Anguillidae), Aplocheilus panchax (Aplocheilidae), Channa sp. (Channidae), Danio aes-
culapii Kullander & Fang, Pethia sp., Puntius chola, Rasbora cf. daniconius (Hamilton),
Rasbora rasbora (Cyprinidae), and Lepidocephalichthys berdmorei (Blyth) (Cobitidae).
Discussion
e rst report of Laubuka laubuca from Bangladesh was by Rahman (1989), who il-
lustrates it with a drawing copied from Day (1878: pl. 151, g. 5), showing a specimen
from Myanmar. e records by Rahman (2005) and Rahman and Chowdhury (2007),
are illustrated with a photo that shows a specimen of Esomus danrica (Hamilton, 1822).
No Laubuka were reported by Ahmed et al. (2013) from their survey of hillstream shes
in northern and southeastern Bangladesh. e distribution of L. laubuca within Bang-
ladesh remains to be mapped. Based on the comparative material used here, Laubuka
laubuca has a wide distribution in the Brahmaputra in Assam, whereas records from
other areas need revision. Laubuka brahmaputraensis Kulabtong, Suksri & Nonpayom,
2012, is the only other species of Laubuka reported from Bangladesh. It was described
on the basis of aquarium specimens without precise collecting locality, but said to be
from the Brahmaputra basin in Bangladesh (Kulabtong et al. 2012), and has been con-
sidered to be a possible junior synonym of L. laubuca (Kottelat 2013). Laubuka tenella
is distinct from L. laubuca in colour pattern and mitochondrial DNA, and there is no
indication that it has been misidentied as L. laubuca in earlier literature.
e Cox’s Bazar and Rakhine samples of L. tenella share the same colour pat-
tern and proportional measurements, but dier slightly in frequencies of anal-n ray
counts. Ranges overlap, however, and the anal n count shows intraspecic variability
also in other species of Laubuka. ere is no variation in dorsal-n count or circumpe-
duncular scale count, but these counts (ii.8½ and 12, respectively) are shared with
most other species in the genus. e miniature species Laubuka fasciata ii.7½ dorsal-
n rays, and 10 circumpeduncular scales. Pethiyagoda et al. (2008) reported dorsal-n
rays ii.9½ in some Sri Lankan species and we observed this count in a specimen of an
unidentied species of Laubuka from Kerala (NRM 12218).

Laubuka tenella, a new species of cyprinid sh from southeastern Bangladesh... 117
Figure 6. Domdomia stream, the type locality of Laubuka tenella, 9 May 2015.
Laubuka khujairokensis, L. latens, L. laubuca, and L. siamensis are characterized by
absence of short dark bars anteriorly on the side. However, Das (1939: g. 1) gured a
specimen identied as L. laubuca from the Damodar River near Hazaribagh (Hugli River
drainage), which shows short indistinct vertical bars on the side, but no lateral stripe.
Otherwise, a series of short vertical bars on the side as in L. tenella, have been reported

Sven O. Kullander et al. / ZooKeys 742: 105–126 (2018)
118
only from Sri Lankan Laubuka, by Pethiyagoda et al. (2008). Laubuka trevori has a nar-
row dark stripe along the side, stated to be less distinct anteriorly (Knight 2015), but ap-
parently no vertical bars. Consequently, L. tenella is diagnosed on the basis of colour pat-
tern and, as veried from Bangladeshi material only, the COI sequence. Laubuka tenella
is also relatively slender, and may dier in that regard from L. laubuca (Fig. 3, Tables 1–2
), and Sri Lankan species (body depth 32.8–34.6 % SL in L. ruhuna, 27.9–32.4 % in
L. varuna, 27.2–30.3 % in L. lankensis according to data in Pethiyagoda et al. (2008).
Laubuka brahmaputraensis was described from three specimens collected by an
aquarium sh collector and said to be from Brahmaputra but without precise locality.
e published photograph (Kulabtong et al. 2012; g. 1) of the holotype, preserved in
1995, suggests that it is in a poor state of preservation, similar to specimens kept long
time in buered formalin. e photo shows a trace of the cleithral spot but no other
markings. e diagnosis and description do not include characters separating from
L. laubuca reported here from Bangladesh and northeastern India. Kulabtong et al.
(2012) did not have specimens of L. laubuca for comparison, but based their concept
of L. laubuca on the meristic data in “Rahman (2003)” which is apparently an error for
Rahman (2005). e validity of this comparison is doubtful.
Rahman’s (2005) description is not in accord with Laubuka from Bangladesh or
northeastern India, but also not compatible with Esomus danrica (Hamilton, 1822),
the species on the illustration of L. laubuca in Rahman (2005). Rahman’s n counts are
compatible with most species of Laubuka, but the counts of 34–36 lateral-line scales,
and 20–21 predorsal scales stand out. e description is almost the same as in Rahman
(1989) and the counts are identical. e high lateral-line counts and predorsal scale
counts seem to be based partly or entirely on earlier literature. Silas (1958) identied L.
laubuca from a wide geographical area, including India, Sri Lanka, Myanmar and adja-
cent south-east Asia, resulting in a lateral line count range of 31–37, as cited by Talwar
and Jhingran (1991). Silas (1958) did not report specimens from Bangladesh, and his
specimens from Myanmar were from east of the Rakhine Yoma. Jayaram (1981, 1999)
stated the lateral line count to be “34 to 37, large, not many”, a count that may go
back to the count of 34–37 in Day (1878: 598). Day’s gure, pl. 151, g. 5, is stated
to show a specimen from Burma and most likely does not show L. laubuca. Day’s de-
scription also covers material from India and Sri Lanka. Day’s Myanmar specimens,
included in L. laubuca by Silas (1958), now in the BMNH, are from Mawlamyine,
Sittaung River, and Mandalay (Silas 1958). Day’s image was copied by later authors
to illustrate L. laubuca, e.g., in Jayaram (1981) and Rahman 1989). Pethiyagoda et al.
(2008) reported elevated counts of 34–37 lateral line scales in L. lankensis and 34–36
in L. insularis, but only and 31–33 in L. varuna and L. ruhuna.
Because earlier concepts of L. laubuca were highly inclusive, including several spe-
cies, and also including information copied from literature sources, published infor-
mation on L. laubuca cannot be relied on for diagnosis of the species. Based on the
locality information provided by Hamilton (1822; Day, 1877), it seems reasonable
that the species of Laubuka present in West Bengal and Assam represents L. laubuca
in the strict sense. is species, with 30–33 lateral-line scales has a colour pattern that

Laubuka tenella, a new species of cyprinid sh from southeastern Bangladesh... 119
includes a distinct vertically oriented cleithral spot, and a thin black stripe from the
middle of the side which is extended caudad but not reaching to the caudal-n base,
and which is slightly expanded on the middle of the caudal peduncle. is colour pat-
tern is illustrated by Pethiyagoda et al. (2008: g. 13), and Lalramliana et al. (2017:
g. 2), and in Fig. 7A–B). Except for the absence of data in Kulabtong et al. (2012),
L.brahmaputraensis agrees with L. laubuca in this strict sense, suggesting that L. brah-
maputraensis is a junior synonym of L. laubuca.
Danio menoni Barman, 1985 from Andhra Pradesh (Barman 1985) was identied
by Tilak and Jain (1987) as a junior synonym of L. laubuca. e image in Barman
(1985) shows clearly that it is a species of Laubuka, but it has a distinct triangular
caudal spot similar to a specimen from Kerala in our comparative material, and may
represent a valid species of Laubuka.
Perilampus perseus M’Clelland [1839: 395, pl. 46, g. 5 (erroneously stated to be
pl. 48, g. 5 in the text]; as P. persus in the index)] was listed as a questionable synonym
of L. laubuca by Day (1878), and has not been recognized as a distinct species since.
e distribution was given as “Assam, and probably Bengal”, and the description of
the only specimen said to be “so much injured that I am unable to determine with ac-
curacy the number of scales and caudal n rays” (M’Clelland, 1839). Knight (2016)
considered M’Clelland’s illustration to be reminiscent of Salmostoma, but particularly
the long pelvic n rather supports the identication as a species of Laubuka. Based on
the locality information by M’Clelland (1839), P. perseus is very likely to be a junior
synonym of L. laubuca. Perilampus guttatus M’Clelland, 1839, is an unneeded replace-
ment name for Cyprinus laubuca Hamilton, 1822 (= L. laubuca).
Figure 7. Laubuka laubuca. A NRM 57234, 53.1 mm SL. India: Brahmaputra River drainage: vicinity
of Dibrugarh B NRM 67317 mm SL, preserved in 95% ethanol. Bangladesh: vicinity of Dhaka.

Sven O. Kullander et al. / ZooKeys 742: 105–126 (2018)
120
Among the Sri Lankan species of Laubuka described by Pethiyagoda et al. (2008),
L. lankensis, and L. insularis were distinguished from L. ruhuna and L. varuna by the
presence of tubercles on the lower jaw. ose structures are, however, neuromast elds,
as observed in L. insularis and other species of Laubuka, and should be expected from
L. varuna and L. ruhuna as well. Specimens identied as possibly L. varuna (LZM
962/5438) have well developed lower jaw neuromast elds. Lower jaw neuromast sen-
sory elds may be a unique character of Laubuka, but have not yet been studied at
organ level. e diagnoses of the Sri Lankan species, based mainly on scale counts,
n lengths and body depth, make it dicult to separate them from continental Asian
congenerics, but it seems unlikely that any of them is conspecic with species in the
northern part of India and Bangladesh.
Günther (1868: 331) described Eustira ceylonensis as a new genus and species based
on six specimens from Ceylon, purchased of Mr Cuming [Hugh Cuming, 1781–
1865]. Day (1878: 599) included E. ceylonensis in Perilampus. Day’s Perilampus equates
Laubuka+Chela. His description of P. ceylonensis is based on data from Günther (1868);
apparently Day had not examined Günther’s specimens. Silas (1957) synonymized E.cey-
lonensis with Danio malabaricus (Jerdon, 1849) (now Devario malabaricus) based on his
own examination of the presumed type series. Silas also provided a line drawing, supplied
by Ethelwynn Trewavas, of a specimen stated to be holotype of E. ceylonensis, and which
seems to show a specimen of Devario. It is noteworthy that the data from the speci-
mens examined by Silas disagree with Günther’s description, while in the same volume
Günther (1868: 282–283) lists specimens of D. micronema (Bleeker, 1863) from Ceylon
(“a, b, c, d-e, f-h. Adult and half-grown”), of which a is stated to have been purchased of
Mr. Cuming; and also material of other species of Danio/Devario. Silas’s assumption that
Günther was unable to distinguish between Laubuka and Devario/Danio may have been
somewhat precipitant. It seems rather that a sample of D. micronema (as identied by
Günther) was mistaken by Silas and Trewavas for the type series of E. ceylonensis.
Günther’s description of E. ceylonensis is compatible with characters of Laubuka.
Günther’s diagnosis of the genus, referring to “entire abdominal edge being trench-
ant”, i.e. keeled, as in Laubuka, stands in contrast to the rounded abdomen in Devario.
e characters “pectorals elongate” (characteristic of Laubuka; not particularly long
in Devario); “barbels none” (as in Laubuka; usually both rostral and maxillary barbels
present in Devario), and “ventrals well developed” (long in Laubuka, not particularly
so in Devario) also suggest a dierent genus than Devario. Günther’s description of the
species E. ceylonensis excludes Devario by reference to absence of a symphyseal knob
of the lower jaw, which is prominent in Devario and Chela, and silvery colour, not
recorded from any Sri Lankan Devario. e species description, however, contains a
statement that is dicult to reconcile with Laubuka or Devario: “Pectoral n shorter
than the head, extending to the ventral.” is statement also seems to be at odds
with the diagnosis of Eustira, which suggests a long pectoral n, even if “elongate”
could also be understood as slender. Other parts of the description of E. ceylonensis
are compatible with both Laubuka and Devario, as well as other cyprinid genera. e
anal-n count of 17 is low for Laubuka, which usually have about 20 anal-n rays, but

Laubuka tenella, a new species of cyprinid sh from southeastern Bangladesh... 121
it may exclude the rst two unbranched rays, which are dicult to discern without
manipulation.
e jar with the syntypes of Eustira ceylonensis could not be located in the BMNH
collection (J. Maclaine, R. Britz, pers. comm.). e specimens incorrectly considered
as the type series of Eustira ceylonensis are catalogued as BMNH 1852.2.19.130–132,
1853.12.24.6, 1864.4–11.33, and are identied by us as tentatively representing De-
vario malabaricus sensu Pethiyagoda (1991). Eustira ceylonensis is probably a senior
synonym of one of the species of Laubuka reported by Pethiyagoda et al. (2008) from
Sri Lanka, but Günther’s description does not contain information that can be used to
identify which of those species is aected.
Although the Myanmar and Bangladeshi samples of Laubuka tenella could not
be separated by morphological characters, it may be noted that meristics and propor-
tional measurements are very conservative in Laubuka. e colour pattern is unique
to L.tenella but similar to that of other species. A longitudinal dark stripe occurs in
several species, although usually narrower than in L. tenella, and short vertical bars
anteriorly on the side is illustrated for L. insularis and L. varuna in Pethiyagoda et al.
(2018). e combination of a posterior dark stripe and anterior series of short verti-
cal bars is unique for L. tenella. e colour pattern is denitely dierent from that
of L. fasciata and L. parafasciata in which there is a horizontal dark stripe along the
middle of the side (Knight 2015: g. 15; Lalramliana et al. 2017: g. 2). Specimens
referable to L. laubuca have plain sides except for a distinct dark cleithral spot, a very
narrow dark stripe posteriorly on the caudal peduncle and, in juveniles only, a dark
spot at the base of the caudal n. In L. siamensis there are indistinct vertical bars
anteriorly on the side, and posteriorly on the side a dark stripe ending in a triangular
blotch on the caudal-n base.
Lalramliana et al. (2017) considered 16 precaudal vertebrae to be diagnostic for
Laubuka parafasciata, contrasting with 14 in other Laubuka. e count of 14 was
probably based on Pethiyagoda et al. (2008) who reported 14 precaudal vertebrae in
Laubuka as diagnostic from Chela cachius with 17. Our specimens of Laubuka possess
15 or 16 vertebrae anterior to that with the rst distinct haemal spine (Table 3). Typi-
cally, vertebrae 5 –14 bear ribs, vertebrae 15 is a short transitional vertebra without ribs
or distinct haemal spine, vertebra 16 presents a haemal spine, and vertebra 17 has a
haemal spine inserted posterior to the ascending rst anal-n pterygiophore (Fig. 8)
ere are no major dierences in vertebral counts between species of Laubuka, as also
suggested by preliminary data from a few species (Table 3).
e terminology of vertebral elements in teleosts is diverse, with confusing syn-
onymy. Günther (1880, 1886) and Hubbs and Lagler (2007; and editions from 1947
onwards) provided the most inuential list of denitions of measurements and counts
in sh taxonomy. Günther distinguished between abdominal and caudal vertebrae,
and Hubbs and Lagler between precaudal and caudal vertebrae, both publications con-
sidering the rst caudal vertebra to be that bearing a denite haemal spine. Günther
(1880) stated more precisely: “the caudal vertebrae dier from the abdominal in hav-
ing the haemapophyseal elements converted into spines similar to the neurals”.

Sven O. Kullander et al. / ZooKeys 742: 105–126 (2018)
122
Naseka (1996) proposed a vertebral column formula recognizing the main divi-
sions of abdominal and caudal vertebrae. Naseka’s abdominal vertebrae include inter-
mediate vertebrae which are vertebrae with parapophyses but no articulating ribs. His
caudal vertebrae include preanal vertebrae with a haemapophysis anterior to the rst
(or rst interhaemal) anal-n pterygiophore, and postanal vertebrae, which are those
posterior to the rst anal-n pterygiophore. Bird and Mabee (2003) mapped vertebral
elements in Danio rerio, distinguishing between Weberian (within the Weberian com-
plex), precaudal (from vertebra 5 to the one before that bearing the rst haemal spine,
“…vertebrae composed of centra, neural arches and spines, parapophyses, and ribs”),
caudal (from that with the rst haemal spine to the one before the caudal-n vertebrae,
“…vertebrae […] composed of centra, neural arches and hemal spines”), and caudal-
n (those supporting the caudal-n rays) vertebrae.
In cypriniforms, there are always four vertebral centra within the Weberian ap-
paratus; posterior vertebrae may vary in number and morphology. In Laubuka the
Weberian vertebrae are followed by trunk vertebrae with long pleural ribs and usu-
ally three vertebrae with no ribs attached. Of these latter three, the anterior two have
parapophyses, a haemal arch, and non-articulating free pleural ribs, whereas the third
is slender and extends ventrad as a haemal spine immediately anterior to the long rst
anal-n pterygiophore or haemapophysis (Fig. 8). On X-radiographs of sh speci-
mens, the shape of the posterior trunk vertebrae commonly cannot be decided, but, as
in Laubuka, their identity is deduced from the relative position, presence or absence
of attached ribs, and the relative length of the haemapophysis/haemal spine. Roberts
(1989), working with X-radiographs, suggested to count as caudal vertebrae those pos-
terior to the rst anal pterygiophore. is method was followed by e.g., Kullander and
Britz (2002), and probably also Lalramliana et al. (2017).
Dierences between authors in reporting vertebral counts can be attributed to dif-
ferent interpretations or denitions of the vertebrae in the transitional zone between
trunk and tail vertebrae. e precaudal count in Pethiyagoda et al. (2008) apparently
Figure 8. X-radiograph of Laubuka tenella. Signicant vertebrae indicated with sequential number.
5=Fifth precaudal vertebra (= 5th abdominal/trunk vertebra); 14 = 14th precaudal vertebra, anterior of two
precaudal vertebrae with haemapophyses and non-articulating ribs/ribs absent (intermediate vertebrae of
Naseka 1996); 15 = last precaudal vertebra; 16 = rst vertebra with haemal spine, which is rst caudal
vertebra of Günther (1880), Naseka (1996), Hubbs and Lagler (2007), and Bird and Mabee (2003), and
the last preanal vertebra of Naseka 1996).

Laubuka tenella, a new species of cyprinid sh from southeastern Bangladesh... 123
refers to the Weberian plus rib-bearing vertebrae, whereas in Lalramliana et al. (2017)
it apparently refers to the Weberian plus remaining trunk vertebrae including the ver-
tebra with the rst haemal spine, i.e., all preanal vertebrae (but not abdominal or pre-
caudal vertebrae as in Günther (1880) or Hubbs and Lagler (2007).
Vertebral numbers and the delimitation of trunk and tail vertebrae are useful taxo-
nomic markers and may explain body proportions. Limitations in resolution of X-ray
images or diculties to determine the rst haemal spine may inuence the classica-
tion of trunk and tail vertebrae, however. It seems relevant in danionin cyprinids to
follow Bird and Mabee (2003) in distinguishing between precaudal vertebrae (i.e. the
ones anterior to the vertebra immediately anterior to the rst anal-n pterygiophore,
understood as bearing the rst haemal spine, but in contrast to Bird and Mabee in-
cluding the four Weberian vertebrae), and caudal vertebrae (all vertebrae posterior to
and including this vertebra, including caudal-n supporting vertebrae, i.e., the the
last half-centrum= compound centrum or preural+ural1) (Fig. 8). is division agrees
with that of Günther (1880, 1886) and Hubbs and Lagler (2007). In groups in which
the shape of transitional trunk vertebrae (Naseka’s intermediate vertebrae) is dicult
to classify, some other method to count or to distinguish the vertebrae may be more
practical. Because the subdivisions are based on convention only, any other method
than those described above are equally valid, but the particular method of counting
vertebrae needs to be explained for every instance to ensure that counts are consistent
and comparisons achievable.
In the phylogenetic analysis, Laubuka tenella came out as sister group to L. laubu-
ca, which has a complementary geographical distribution west of L. tenella. Only few
sequences of Laubuka were available from GenBank to supplement our material from
India and Bangladesh, and among them several lacked locality information (Fig. 4). In-
corporation of additional taxa, e.g., from Sri Lanka, Myanmar and Indochina may al-
ter signicantly the pattern shown in Fig. 4. Nevertheless, the position in the tree (Fig.
4) and the 9 % p-distance dierence from the most similar species, are strong indica-
tors of species distinctness of L. tenella. One of the downloaded GenBank sequences,
KT353103, from northern Bangladesh, and identied as L. laubuca in GenBank, is
resolved in the L. laubuca clade (Fig. 4), but as sister to the remaining specimens. A
possible explanation is that it represents one more species of Laubuka in the region, but
could also reect wide genetic variation in L. laubuca.
Laubuka tenella is known so far only from three localities, and DNA data are available
from only two localities, representing two distinct haplotypes. It seems plausible to con-
Table 3. Vertebral counts in Laubuka tenella and comparative material of Laubuka.
Species N 14+19=33 15+18=33 15+19=34 15+20=35 16+18=34 16+19=35
L. tenella 13 2 9 2
L. parafasciata 6 4 1 1
L. laubuca 2 1 1
L. insularis 2 2
L. siamensis 5   1 4  

Sven O. Kullander et al. / ZooKeys 742: 105–126 (2018)
124
sider the localities to be part of a continuous distribution, but the dierences in the COI
sequence between the two Bangladeshi localities may be an indication of fragmentation.
e coasts of extreme southeastern Bangladesh and southwestern Myanmar are ichthyo-
logically underexplored. e Rakhine Yoma is a hill range separating the sh fauna in the
lowlands to the west from that of the Irrawaddy basin, exemplied by several endemic
species from western Rakhine Yoma (e.g., Kullander 2015, Kullander and Britz 2015).
is lower costal region extends into southeastern Bangladesh, and it is not surprising to
nd the same species both in southwestern Myanmar and southeast, although L. tenella
may be the rst species reported as endemic for this particular region. Devario anomalus,
in the Cox’s Bazar district, and D. xyrops Fang & Kullander, 2009, on the Western slope
of the Rakhine Yoma, are two very similar sister taxa, which are considered as distinct spe-
cies (Kullander et al. 2017), but with a combined distribution similar to that of D. tenella.
Acknowledgements
Specimens were collected with the assistance of Abdullah Helal, Mozammel Hossain,
Suman Mandal, and Mamunur Rashid, University of Dhaka. Research on Bangladeshi
freshwater shes was supported by the project “Genetic characterization of freshwater
shes in Bangladesh using DNA barcodes” funded by the Swedish Research Council,
contract D0674001 to Sven Kullander and Abdur Rob Mollah. Specimens of Laubuka
parafasciata were put at our disposal by Ralf Britz, who also provided clarifying com-
ments on morphology of vertebrae. James Maclaine and Ralf Britz searched for the
types of Eustira ceylonensis.
References
Ahmed ATA, Rahman MM, Mandal S (2013) Biodiversity of hillstream shes in Bangladesh.
Zootaxa 3700: 283–293. https://doi.org/10.11646/zootaxa.3700.2.6
Arunkumar L (2000) Fishes of the genus Chela Hamilton - Buchanan (Cyprinidae: Cultrinae)
from Manipur, India, with description of a new species. Aquaculture 1: 121–124.
Banarescu P (1968) Remarks on the genus Chela Hamilton-Buchanan (Pisces, Cyprinidae)
with description of a new subgenus. Annali del Museo Civico di Storia Nturale “Giacomo
Doria” 78: 53–63.
Barman RP (1985) A new cyprinid sh of the genus Danio Hamilton (Pisces: Cyprinidae) from
Andhra Pradesh, India. Journal of the Bombay Natural History Society 82: 602–605.
Bird NC, Mabee PM (2003) Developmental morphology of the axial skeleton of the zebrash,
Danio rerio (Ostariophysi: Cyprinidae). Developmental Dynamics 228: 337–357. https://
doi.org/10.1002/dvdy.10387
Conway KW, Mayden RL, Tang KL (2009) Devario anomalus, a new species of freshwater sh
from Bangladesh (Ostariophysi: Cyprinidae). Zootaxa 2136: 49–58.
Day F (1875–1878) e shes of India; being a natural history of the shes known to inhabit the
seas and fresh waters of India, Burma, and Ceylon. Quaritsch, London 1875: 1–168. [pls.

Laubuka tenella, a new species of cyprinid sh from southeastern Bangladesh... 125
1–40; 1876: 169–368, pls. 41–78; 1877: 369–552, pls. 79–138; 1878: i–xx, 553–778, pls.
139–195]
Day F (1877) e sh and sheries of Bengal. In: Hunter WW (Ed.) A statistical account of
Bengal Volume XX. Trübner & Co., London, 1–120.
Fang F (1997) Redescription of Danio kakhienensis, a poorly known cyprinid sh from the
Irrawaddy basin. Ichthyological Exploration of Freshwaters 7: 289–298.
Günther A (1868) Catalogue of the shes in the British Museum. Volume seventh. Trustees of
the British Museum, London, 512 pp.
Günther ACLG (1880) An introduction to the study of shes. Adam and Charles Black,
Edinburgh, 720 pp. http://dx.doi.org/10.5962/bhl.title.54205
Günther ACLG (1886) Handbuch der Ichthyologie. Carl Gerold’s Sohn, Wien, 527 pp.
Hamilton, F (1822) An account of the shes found in the River Ganges & its branches. Archibald
Constable and Company, Edinburgh, 405 pp. [39 pls]
Hubbs CL, Lagler KF (revised by G.R. Smith) (2007). Fishes of the Great Lakes Region. Revised
edition, e University of Michigan Press, Ann Arbor, 276 + 5 pp. [32 pls]
Huelsenbeck, JP, Ronquist F (2001) MRBAYES: Bayesian inference of phylogenetic trees.
Bioinformatics 18: 754–755. https://doi.org/10.1093/bioinformatics/17.8.754
Jayaram KC (1981) e freshwater shes of India. A handbook. Zoological Survey of India,
Calcutta, 475 pp. [12 pls]
Jayaram KC (1999) e freshwater shes of the Indian region. Narendra Publishing House,
Delhi, 551 pp.
Kearse M, Moir R, Wilson A, Stones-Havas, S, Cheung M, Sturrock S, et al. (2012) Geneious
Basic: an integrated and extendable desktop software platform for the organization and
analysis of sequence data. Bioinformatics28: 1647–1649. https://doi.org/10.1093/bioin-
formatics/bts199
Knight JDM, Rema Devi K (2014) Chela macrolepis a new species of cyprinid sh from south-
ern India (Teleostei: Cyprinidae). Ichthyological Exploration of Freshwaters 25: 156–166.
Knight JDM (2015) Description of two new species of Laubuka (Teleostei: Cyprinidae) from River
Cauvery, southern India. Zootaxa 4000: 518–530. https://doi.org/10.11646/zootaxa.4000.5.2
Knight JDM, Rema Devi K (2014) Chela macrolepis, a new species of cyprinid sh from south-
ern India (Teleostei: Cyprinidae). Ichthyological Exploration of Freshwaters 25: 159–166.
Kottelat, M. (2013) e shes of the inland waters of Southeast Asia: a catalogue and core
bibliography of the shes known to occur in freshwaters, mangroves and estuaries. e
Raes Bulletin of Southeast Asian Zoology, Supplement 27: 1–663.
Kulabtong S, Suksri S, Nonpayom C (2012) A new species of genus Laubuca Bleeker, 1860
cyprinid sh from Bangladesh (Cypriniformes, Cyprinidae). Biodiversity Journal 3: 93–95.
Kullander SO (2015) Taxonomy of chain Danio, an Indo-Myanmar species assemblage, with
descriptions of four new species (Teleostei: Cyprinidae: Danioninae). Ichthyological Explora-
tion of Freshwaters 25: 355–378.
Kullander SO (2017) Devario fangae and Devario myitkyinae, two new species of danionine
cyprinids from northern Myanmar (Teleostei: Cyprinidae: Danioninae). Zootaxa 4227:
407–421. https://doi.org/10.11646/zootaxa.4227.3.7
Kullander SO, Britz R (2002) Revision of the family Badidae (Teleostei: Perciformes), with descrip-
tion of a new genus and ten new species. Ichthyological Exploration of freshwaters 13: 295–372.

Sven O. Kullander et al. / ZooKeys 742: 105–126 (2018)
126
Kullander SO, Britz, R (2015) Description of Danio absconditus, new species, and redescription
of Danio feegradei (Teleostei: Cyprinidae), from the Rakhine Yoma hotspot in south-west-
ern Myanmar. Zootaxa 3948: 233–247. http://dx.doi.org/10.11646/zootaxa.3948.2.5
Kullander SO, Rahman MM, Norén M, Mollah AR (2017) Devario in Bangladesh: Species diver-
sity, sibling species, and introgression within danionin cyprinids (Teleostei: Cyprinidae: Dan-
ioninae). PLoS ONE 12(11): e0186895. https://doi.org/10.1371/journal.pone.0186895
Lalramliana D, Valalhlimpuia, Singh M (2017) Laubuka parafasciata, a new cyprinid sh spe-
cies (Teleostei: Cyprinidae) from Mizoram northeastern India. Zootaxa 4244: 269–276.
https://doi.org/10.11646/zootaxa.4244.2.8
M’Clelland J (1839) Indian Cyprinidae. Asiatic Researches 19: 217–471 + 1.
Masters BC, Fan V, Ross HA (2011) Species delimitation – a Geneious plugin for the exploration
of species boundaries. Molecular Ecology Resources 11: 154 –157. https://doi.org/10.1111/
j.1755-0998.2010.02896.x
Naseka AM (1996) Comparative study on the vertebral column in the Gobioninae (Cyprinidae,
Pisces) with special reference to its systematics. Publicaciones Especiales, Instituto Español
de Oceanografía 21: 149–167.
Pethiyagoda R, Kottelat M, Silva A, Maduwage K, Meegaskumbura M (2008) A review of the
genus Laubuca in Sri Lanka, with description of three new species (Teleostei: Cyprinidae).
Ichthyological Exploration of Freshwaters 19: 7–26.
Rahman AKA (1989) Freshwater shes of Bangladesh. Zoological Society of Bangladesh, Dha-
ka, 364 pp.
Rahman AKA (2005) Freshwater shes of Bangladesh. Second edition. Zoological Society of
Bangladesh Dhaka, 394 pp.
Rahman AKA, Chowdhury GW (2007) 46. Chela laubuca (Hamilton, 1822). In: Siddiqui KU,
Islam MA, Kabir SMH, Ahmad M, Ahmed ATA, Rahman AKA, Haque EU, Ahmed ZU,
Begum ZNT, Hassan MA, Khondker M and Rahman MM (Eds) Encyclopedia of Flora and
Fauna of Bangladesh, vol. 23, Freshwater Fishes. Asiatic Society of Bangladesh, Dhaka, 50.
Rambaut A, Suchard M, Drummond A (2014) Tracer. http://treebioedacuk/software/tracer/
Roberts TR (1989) e freshwater shes of Western Borneo (Kalimantan Barat, Indonesia).
Memoirs of the California Academy of Sciences 14: 1–210.
Ronquist F, Huelsenbeck JP (2003) MrBayes 3: Bayesian phylogenetic inference under mixed
models. Bioinformatics19: 1572–1574. https://doi.org/10.1093/bioinformatics/btg180
Silas, EG (1957) e Ceylonese cyprinid genus Eustira Günther considered a synonym of
Danio Hamilton. Copeia 157: 61–62. https://doi.org/10.2307/1440532
Silas EG (1958) Studies on cyprinid shes of the Oriental genus Chela Hamilton. Journal of the
Bombay Natural History Society 55: 54–99.
Smith HM (1931) Descriptions of new genera and species of Siamese shes. Proceedings of the
United States National Museum 79(7): 1–48. https://doi.org/10.5479/si.00963801.79-2873.1
Srivathsan A, Meier R (2011) On the inappropriate use of Kimura‐2‐parameter (K2P) diver-
gences in the DNA‐barcoding literature. Cladistics 28: 190–94. http://dx.doi.org/10.1111/
j.1096-0031.2011.00370.x
Systat Software (2009) Systat version 13. Systat Software, Inc., San Jose.
Talwar PK, Jhingran AG (1991) Inland shes of India and adjacent countries, Volume 1. Ox-
ford & IHB Publishing Co, Pvt. Ltd, New Delhi, Bombay & Calcutta, 541 pp.