Labeo heladiva, Sudasinghe & Ranasinghe & Goonatilake & Meegaskumbura, 2018

Sudasinghe, Hiranya, Ranasinghe, R. H. Tharindu, Goonatilake, Sampath Alwis & Meegaskumbura, Madhava, 2018, A review of the genus Labeo (Teleostei: Cyprinidae) in Sri Lanka, Zootaxa 4486 (3), pp. 201-235 : 204-214

Labeo heladiva , new species

( Figures 2–3 )

Labeo dussumieri (from Sri Lanka, not Valenciennes, 1842): Günther, 1868: 59; Day, 1889: 262; Duncker, 1912: 261; Deraniyagala, 1952: 41; Munro, 1955: 46; Mendis & Fernando, 1962: 117; Senanayake, 1980: 146; Pethiyagoda, 1991:

82; Talwar & Jhingran, 1991: 206; Jayaram & Dhas, 2000: 22.

Material examined. Holotype: 2018.08.01.NH, 134 mm SL, Sri Lanka, Attanagalu Oya basin: Uruwal Oya, 7°03'09"N 80°03'09"E, 17 m asl. H. Sudasinghe and R.H.T. Ranasinghe. Apr 2017.

……continued on the next page ……continued on the next page

Paratypes: DZ 3821, DZ 3148, DZ 3166, DZ 3168, 7, 108– 122 mm SL, Sri Lanka, Mahaweli River basin, Polonnaruwa: Amban river, 7°51'53"N 80°59'29"E, 54 m asl. H. Sudasinghe. Jun 2015; UPZM uncatalogued, 10, 201– 257 mm SL, Sri Lanka .

Other material: Identified but not included in morphometric data. WHT 7904, 70.7 mm SL, Sri Lanka, Kala Oya basin, Eluwankulama; WHT 9258, 21, 73.9–117 mm SL, Sri Lanka, Mahaweli River basin, Polonnaruwa ; WHT 30837, 133 mm SL, Sri Lanka, Mahaweli River basin, Seruwila ; WHT 30863, 277 mm SL, Sri Lanka, Mahaweli River basin, Wasgamuwa.

Diagnosis. Labeo heladiva is distinguished from all Sri Lankan and peninsular-Indian congeners by the combination of the following characters: two pairs of barbels (maxillary and rostral); dorsal fin with 12–13 branched rays; lateral line with 44–51 scales; scales in transverse series ½8–½9+1+6–7; circumpeduncular scales 19–22; eye diameter 18.4–24.4 % HL; in life, 6–7 rows of scales on side of body above and between tip of pectoral fin and origin of anal fin with orange-colored patches, giving the appearance of an orange blotch; upper and lower margins of scales on the side of the body dark-pigmented, forming 9–13 hazy lines.

Description. For general appearance, see Figure 2 ; morphometric data are provided in Table 3. Maximum size 277 mm SL. Caudal peduncle short, its depth 73.4–97.1 % its length. Eyes medium-sized (18.4–21.5 % HL in specimens> 200 mm SL, 21.6–24.4 % HL in specimens <200 mm SL), located dorsolaterally, but visible in ventral view.

Rostral fold poorly developed, slightly overlapping upper lip. Tubercles conoid, tuberculation prominent on rhinal, rostral, infraorbital fields; minute tuberculation on preorbital field ( Figure 3 ). Upper labial fold with prominent lobed papillae in 4–5 rows; lower labial fold with prominent lobed papillae in 2–3 rows ( Fig. 3C ). Maxillary barbel embedded in lip fold, tip externally visible. Rostral barbel located on medial margin of rostral flap. Maxillary barbel longer than rostral, just reaching vertical through anterior-most point of nares.

Dorsal fin with two simple and 12 (9) or 13 (4) branched rays; first simple ray stiff, about half length of second. Pectoral fin with one simple and 14 (2), 15 (6) or 16 (6) branched rays. Pelvic fin with one simple and eight (14) branched rays. Origin of pelvic fin beneath 4th branched ray of dorsal fin. Anal fin with two simple and five (13) branched rays; first simple ray stiff, less than half length of second simple ray. Caudal fin forked, with 9+9 (2) or 9+8 (9) branched rays in upper and lower lobe, respectively. Upper caudal-fin lobe slightly longer than lower.

Lateral line complete, with 44 (2), 45 (6), 46 (6), 47 (8), 48 (8), 49 (6), 50 (4) or 51 (2) + 1–3 scales. Scales in transverse series ½8+1+6 (11), ½8+1+6½ (6), ½8+1+7 (7), ½8+1+7½ (1), 9+1+6 (5), 9+1+6½ (3), 9+1+7 (2), ½9+1+6½ (1) or ½9+1+7 (5). Predorsal scales 19 (1), 20 (4), 21 (11) or 22 (2). Prepelvic scales 30 (1), 32 (1), 33 (1), 34 (2), 35 (2), 36 (1), 37 (3), 38 (3). Circumpeduncular scales 19 (1), 20 (10), 21 (24), 22 (7).

Coloration. In 70% alcohol ( Fig. 2B ): head and body dull greyish brown dorsally, becoming lighter laterally, off-white ventrally. All fins with greyish-black melanophores along rays. Interradial membrane of fins darker than rays. Tubercles white. Black blotch at caudal peduncle about seven scales long, five high. Lateral body with 9–13 hazy black lines, extending from opercular membrane to caudal peduncle (these lines result from the fusion of black pigments on the dorsal and ventral scale margins).

In life ( Fig. 2A ): dorsally silvery grey, becoming lighter laterally. Venter white. Six to seven rows of scales on sides of body above and between distal margin of pectoral fin and origin of anal fin with orange-colored patches, giving the appearance of a vague orange blotch. Black blotch at caudal peduncle, about seven scales long, five scales high. Dorsal, caudal, pectoral fins dull greyish-brown to hyaline. Pelvic and anal fins lighter (or hyaline), with dark pigmentation at their bases. Tubercles whitish. Side of body with 9–13, hazy black lines, extending from opercular membrane to caudal peduncle, along dorsal and ventral margin of scales.

Etymology. The species name, heladiva , is a historical Sinhala name for Sri Lanka; applied as a noun in apposition.

Comparative morphometrics. Labeo heladiva is distinguished from L. fisheri by having 44–51 (vs. 37–39: Fig. 4A ) lateral line scales; a count of ½8–½9+1+6–7½ (vs. 7–8+1+4½–6: Fig. 4B ) scales in transverse series; 19– 22 (vs. 17–20: Fig. 4D ) circumpeduncular scales; and 2–3 (vs. 4–5) rows of prominent lobed papillae on the lower labial fold. It differs from the only other species of Labeo native to Sri Lanka , L. lankae, by having a greater lateralline scale count (44–51 vs. 36–39: Fig. 4A ); and rows of 2–3 (vs. 4–5) prominent lobed papillae on the lower labial fold.

Labeo heladiva can be distinguished from L. rohita (now naturalized in Sri Lanka), by having 44–51 (vs. 38– 40) lateral-line scales and ½8–½9 (vs. ½6–½7) scales between the dorsal-fin origin and the lateral line.

Labeo heladiva can be distinguished from the South Indian L. dussumieri , its closest congener, by possessing longer rostral barbels (5.5–10.2% HL vs. 4.0–5.1%); 44–51 vs. 50–60 lateral-line scales; 19–22 vs. 22–27 circumpeduncular scales; and 5–6½ vs 6½–7½ scales from the lateral line to the anal-fin origin.

Seven other peninsular-Indian species of Labeo are considered to be valid (Eschmeyer et al. 2017): Labeo potail (Sykes) , L. porcellus (Heckel) , L. boggut (Sykes) , L. fimbriatus (Bloch) , L. kawrus (Sykes) , L. kontius (Jerdon) and L. nigrescens Day. In addition, Jayaram & Dhas (2000) recorded L. boga (Hamilton) , L. calbasu (Hamilton) , L. dyocheilus (McClelland) , L. gonius (Hamilton) and L. pangusia (Hamilton) from peninsular India.

Labeo heladiva differs from L. porcellus , L. potail , L. boggut , L. kawrus , L. kontius , L. nigrescens , L. boga , L. gonius and L. pangusia by having 44–51 lateral-line scales (vs. 36–37 in L. porcellus ; 39–41 in L. potail ; 55–65 in L. boggut ; 38 in L. kawrus ; 30–42 in L. kontius ; 36–37 in L. nigrescens ; 37–39 in L. boga ; 65–80 in L. gonius ; and 40–42 in L. pangusia ). Labeo heladiva further differs from L. boggut , L. kawrus , L. boga , L. dyocheilus , L. pangusia , and L. potail by having two pairs of barbels (vs. a single pair, except in L. potail , which altogether lacks barbels). The new species can be distinguished from L. porcellus , L. boggut , L. fimbriatus , L. nigrescens , L. boga , L. calbasu , and L. gonius by having ½8–½9+1+6–7½ scales in transverse series (vs. ½6–½7+1+½ 5–6 in L. porcellus ; ½10–12+1+8½– 9 in L. boggut ; ½9–10+1+6½–7½ in L. fimbriatus ; ½6–½7+1+5½ in L. nigrescens ; 7– ½7+1+5–5½ in L. boga ; ½7–½9+1+5½–6½ in L. calbasu ; 12–14+1+10½– 13 in L. gonius ). Further, L. heladiva differs from L. boggut , L. fimbriatus ; L. kawrus , L. nigrescens ; L. boga ; L. dyocheilus ; and L. gonius by having 12– 13 branched dorsal-fin rays (vs. 9–10 in L. boggut ; 15–19 in L. fimbriatus ; 9 in L. kawrus ; 14 in L. nigrescens ; 8–9 in L. boga ; 9–10 in L. dyocheilus ; and 14–16 in L. gonius ).

Reconstruction of haplotype network for Labeo dussumieri and L. heladiva . The two TCS networks for the COI and cytb genes formed two clearly-separated haplotype groups for the Sri Lankan L. heladiva and the Indian L. dussumieri , with no sharing of haplotypes between the two species ( Figure 5 ). In Sri Lanka, the wet zone (H 1 in the two networks) and dry zone populations (H2 and H3) of L. heladiva formed two distinct groups with 5 and 15 mutations for COI and cytb, respectively. The Indian L. dussumieri is more divergent from the Sri Lanka wet-zone population of L. heladiva than the Sri Lankan dry zone population for the cytb genes (25 mutations, vs. 20). However, for COI, the Indian population of L. dussumieri showed a greater divergence from the Sri Lankan dry zone population of L. heladiva than the wet zone (10 vs 7 mutations). Labeo rajasthanicus Datta & Majumdar , a species recently validated by Lal et al. (2015) is morphologically and genetically similar to L. dussumieri . However, L. rajasthanicus is divergent from L. heladiva by a minimum of 10 mutations for the COI gene.

Figures

FIGURE 2. A–B, Coloration of Labeo heladiva A, in aquarium; B, in preservation, holotype 2018.08.01.NH, 134 mm SL, Sri Lanka; and C, L. dussumieri, NBFGRKU uncatalogued, 255 mm SL, India.

FIGURE 3. Detail of head of Labeo heladiva and L. dussumieri in A,D, lateral; B,E, dorsal; and C,F, ventral view. (A–C, L. heladiva, holotype, 2018.08.01.NH, 134 mm SL, Sri Lanka; D–F, L. dussumieri, NBFGRKU uncatalogued, 255 mm SL, India).

FIGURE 4. Boxplots representing the distribution of A, lateral-line scales; B, total scales in transverse series; C, scales between lateral line and anal-fin origin; and D, circumpeduncular scales in Labeo lankae (N=9), L. porcellus (N=3), L. fisheri (N=14), L. heladiva (N=42), and L. dussumieri (N=7), based on material examined.

FIGURE 5. The TCS Haplotype network for Labeo heladiva (red) and L. dussumieri (green) populations based on the analysis of A, 666 bp fragment of the COI; and B, 1074bp fragment of the cytb genes. The sizes of the circles are proportional to the number of individuals sharing a given haplotype. The number of mutational steps are indicated by hatch marks. The black circles are hypothetical nodes.