Gonorhynchus McClelland 1838

(Figures 1, 2, 3 A, 3C, 5, 7, 8, 11)

Gonorhynchus McClelland 1838:943 (type species: Gonorhynchus brevis M’Clelland 1839:373, by subsequent designation by Jordan 1919:195). Gender masculine.

Tariqilabeo Mirza & Saboohi 1990:405 (subgenus of Labeo Cuvier, 1816:194; type species: Labeo macmahoni Zugmayer 1912:597, by original designation). Gender masculine.

Akrokolioplax Zhang & Kottelat 2006:21 (type species: Epalzeorhynchos bicornis Wu 1977, by original designation). Gender feminine.

McClelland (1838) first described the genus Gonorhynchus, but without any discussion of inclusive species or type designation. Jordan (1919) subsequently designated Gonorhynchus brevis M’Clelland 1839 as the type species, now recognized as a junior synonym of G. latius (Hamilton 1822). In the same publication, however, Jordan (1919) synonymized Gonorhynchus with Crossocheilus as M’Clelland (1839) had needlessly renamed Cyprinus gohama Hamilton 1822 (which at the time of Jordan’s publication was considered a junior synonym of Crossocheilus latius) as G. b re vi s.

Cyprinus gohama was initially considered a junior synonym of Cyprinus latius Hamilton 1822 (= Crossocheilus latius) by Day (1877) but placed under the name Cirrhina (= Cirrhinus Oken 1817) latia. Curiously, Day (1878) later synonymized several species, including Crossocheilus latius (Hamilton 1822), under Cirrhina gohama (Hamilton 1822), but later Day (1880) reverted back to the name Cirrhina latia. Cyprinus sada Hamilton 1822 was consistently treated as a junior synonym in these publications (Day 1877, 1878). Unfortunately, the type specimens of none of these nominal taxa are available. Specimens from the Koshi River (eastern Nepal into Bihar, India), one of the rivers from which Cyprinus gohama was originally described, were examined by us, but are not diagnosable from specimens from the Teesta River, the type locality of C. latius (see account for G. latius below), based on the characters listed by Hamilton (1822) or additional characters examined here. As specimens from the original type localities are indistinguishable and Day (1877) served as first reviser, precedence is given to G. latius as the senior synonym of the type species of the genus, G. brevis.

The similarly named genus Gonorynchus Scopoli 1777 is comprised of marine species in the order Gonorynchiformes. Gonorynchus has been misspelled as “ Gonorhynchus ” in a few instances, as noted by Grande (1999).

Diagnosis. Gonorhynchus (Fig. 2) belongs to the tribe Labeonini based on the following characters: 1) incised border on the first anal-fin pterygiophore; 2) ventrally expanded rostral fold; and 3) presence of a superficial posterior labial fold (Stiassny & Getahun 2007). Reid (1982) diagnosed Labeonini as having a vomero-palatine organ, a terete process on the basioccipital, and the neural complex of the Weberian apparatus in direct contact with the supraoccipital region. Although these characters could not be examined in specimens on loan to us, Stiassny & Getahun (2007) reported them as present in G. diplochilus. Molecular phylogenies place species of Gonorhynchus within Labeonini (Wang et al. 2007; Zheng et al. 2010, 2012; Yang et al. 2012).

Gonorhynchus is distinguished from other labeonin genera (see Yang et al. [2012] for list of genera) by the following combination of characters: 1) rostral cap covering upper lip, which is not visibly separated from snout by groove and not attached to lower lip; 2) upper lip very thin or absent, widening substantially at corner of mouth connecting upper jaw to lower lip; 3) inferior mouth; 4) lower lip free only on anterior and lateral edges, posterior edge connected to underside of head (not modified into rounded mental disc), central region equally thick as lateral edges, anterior edge with large papillae; 5) sublachrymal groove uniformly narrow from corner of mouth to rostral barbel, not expanding anteriorly to contain rostral lobe (rostral lobe absent); 6) 8½ branched dorsal rays.

Species included. Gonorhynchus latius (Hamilton 1822), G. diplochilus (Heckel 1838), G. wattanah (Sykes 1839), G. macmahoni (Zugmayer 1912), G. burmanicus (Hora 1936), G. b i c o r n i s (Wu 1977), and G. periyarensis (Menon & Jacob 1996).

Comparison. Gonorhynchus is herein compared with morphologically similar labeonins from Asia. Gonorhynchus is distinguished from Crossocheilus by the absence of a rostral lobe (although a rostral flap may be present) (Fig. 3 A) vs. present (Fig. 3 B) and a very thin upper lip (sometimes absent) that widens at the corner of the mouth (Fig. 3 C) vs. a thicker upper lip that does not widen at the corner of the mouth (Fig. 3 D). In the material examined, Gonorhynchus species have generally more vertebrae (34–38 vs. 31–33) and pored lateral scales (lateral-line scales plus pored scales on caudal fin 35–42 vs. 28–35) than Crossocheilus species. These higher counts in Gonorhynchus are consistent with data from previous descriptions, including those of Heckel (1838), Day (1877), Zugmayer (1912), Menon & Jacob (1996), and Zhang & Kottelat (2006). Kullander et al. (1999) listed several characters differentiating Gonorhynchus and Crossocheilus based primarily on examinations of G. diplochilus and C. oblongus. Two of these characters are listed above (e.g., relative upper lip size and presence/ absence of rostral lobes), but the remaining characters were not found to distinguish the two genera upon examination of additional species.

Gonorhynchus differs from Garra through the absence of a mental sucking disc. Epalzeorhynchos possesses movable rostral lobes just dorsal to the rostral barbels, which are absent in Gonorhynchus, although G. b i c o r n i s has movable flaps on the tip of the snout. Differences between G. b i c o r n i s and several other labeonin genera (see below) discussed by Zhang & Kottelat (2006) apply to all Gonorhynchus species and these genera. Unlike Paracrossochilus, Gonorhynchus has a vestigial and non-papillose upper lip, if lip is present (Kottelat et al. 1993; Zhang & Kottelat 2006). Gonorhynchus differs from Rectoris and Pseudocrossocheilus by having the lower lip disconnected from the rostral cap (vs. connected), and from Sinocrossocheilus in the presence of 8½ (vs. 7½) branched dorsal rays, the upper jaw connected (vs. disconnected) to the lower lip by a frenum around the corner of the mouth, and the absence (vs. presence) of a centrally protruding lower lip covered by papillae (Zhang & Chen 2004; Zhang & Kottelat 2006; Yuan et al. 2008; Zhu et al. 2012).

Remarks. As noted by Kullander et al. (1999), G. macmahoni is morphologically similar to G. diplochilus based on initial descriptions (Zugmayer 1912; 1913). In an examination of freshly collected specimens, Mirza & Saboohi (1990) noted the presence of lateral lobes on the snout of G. macmahoni, which would exclude this species from Gonorhynchus as diagnosed here. However, without an examination of these specimens (whereabouts unknown), it is unclear as to the exact location and features of these structures. The status of this species was most recently reviewed again by (Mirza & Arshard 2008), who observed the syntype to be morphologically similar to G. diplochilus, except in lacking fringes present on the rostral fold in G. diplochilus as well as all other species of Gonorhynchus. Additionally, there was no mention of the presence of rostral lobes by Mirza & Arshard (2008). We examined digital images of the syntype of G. macmahoni (NMW 81256), discussed and shown in figure 1 (page 465) of Mirza and Arshard (2008), and observed no rostral lobes. Also, there appear to be no fringes on the rostral fold, which distinguishes this species from all other species in the genus, which possess well-developed fringes, or fimbriae, on the rostral fold. We tentatively retain G. macmahoni as a valid species pending the examination of additional specimens from the type locality, the Dasht River, a coastal river in western Pakistan draining directly into the Arabian Sea, and surrounding areas.

Gonorhynchus adiscus was revalidated by Sayyadzadeh et al. (2015) based on morphological data and a phylogenetic analysis of the mitochondrial cytochrome c oxidase I (COI) gene. This species was previously synonymized with G. diplochilus by Bianco & Bănărescu (1982). Characters used by Sayyadzadeh et al. (2015) to diagnose G. a d i s c u s and G. diplochilus included the length of rostral barbels, the presence and length of maxillary barbels, counts of rakers on the first gill arch, and counts of scales between the anus and anal-fin insertion, based on specimens from the Sistan, Mashkid, and Makran basins in eastern Iran and western Pakistan (Table 1). However, in their morphological comparisons, Sayyadzadeh et al. (2015) did not include specimens from the Indus River where G. diplochilus is extensively distributed in the headwaters, or any from the Helmand River in Afghanistan where G. a d i s c u s is also reported. An examination of Gonorhynchus specimens from the Indus and Helmand Rivers reveals that the characters listed by Sayyadzadeh et al. (2015) do not differentiate G. a d i s c u s from G. diplochilus.

Table 1 contains comparisons of characters between Sayyadzadeh et al. (2015) from specimens in the Sistan, Makran, and Mashkid basins and specimens examined here from the Helmand and Indus Rivers. The inclusion of specimens from the Helmand and Indus Rivers indicates substantial overlap in the lengths of the rostral and maxillary barbels, the number of rakers on the first gill arch, and the number of scales between the anus and analfin insertion between populations re-validated as G. a d i s cu s (Sistan and Helmand basins) and those treated as G. diplochilus (Makran, Mashkid, and Indus River basins) by Sayyadzadeh et al. (2015). Sharma et al. (2014) also reported gill raker counts of 17–21 in specimens of G. diplochilus from the Poonch River (Indus River basin) in India, which overlap those of G. adiscus reported by Sayyadzadeh et al. (2015). An additional specimen examined by us from the Makran region of southeastern Iran (USNM 205897) was consistent with data presented for G. diplochilus in the same region by Sayyadzadeh et al. (2015) in lacking maxillary barbels, possessing a rostral barbel length equal to 11.7% of head length, and possessing two scales between the anus and anal-fin insertion (we were unable to get a count of rakers on the first gill arch). Based on these additional morphological data, we recommend that G. adiscus continue to be treated as a junior synonym of G. diplochilus. The molecular phylogeny presented in Sayyadzadeh et al. (2015) resolved G. adiscus and G. diplochilus as monophyletic clades, with G. adiscus more closely related to G. latius than to G. diplochilus, supporting their claim for these populations as separate species. However, this phylogeny may not accurately reflect species relationships as only the COI gene was incorporated. The broad geographic range of G. diplochilus from India to Iran holds the possibility of multiple opportunities for divergence via allopatric speciation; however, thorough sampling of populations from throughout this range is required to collect additional morphological and molecular data to test for population differences that would lead to species diagnoses.

The description, illustration, and reported similarities to other Gonorhynchus species of the specimen from the Brahmaputra basin used in the description of G. gobioides in M’Clelland (1839) suggest it is morphologically similar to G. latius. Based on M’Clelland (1839), G. gobioides is distinguished from all other species in the genus by the absence of barbels as all other species in the genus possess at least one pair of barbels on the snout. Bleeker (1860) and Günther (1868) recognized the species as valid in the genus Crossocheilos. Curiously, no mention is made of G. gobioides in the revisions of Indian fishes by Day (1877, 1878) or in the check list of freshwater fishes of India (Menon 1999). M’Clelland (1839) noted the absence of barbels in another species—therein G. petrophilus, currently recognized as a junior synonym of Schizothorax richardsonii (Gray 1832), which possesses two pairs of barbels (Kullander et al. 1999; Chen & Chao 2000; He & Chen 2006). It is possible that M’Clelland (1839) also missed this character in the description of G. gobioides, particularly if dealing with a small or damaged specimen. Furthermore, as noted by M’Clelland (1839), the specimen of G. gobioides was questionably the same species described as Cyprinus mosario by Hamilton (1822). The description of C. mosario lacks details, with the species being described as similar to G. latius except in lacking “tendrils” (= barbels). We have observed no specimens of Gonorhynchus lacking rostral barbels. In light of the absence of the types for both species to confirm the absence of barbels, but their otherwise similarity to G. latius, we recognize G. gobioides and C. mosario as questionable junior synonyms of G. latius. The remaining species of Gonorhynchus discussed in M’Clelland (1839) that have not been synonymized with G. latius (G. bimaculatus, G. brachypterus, and G. caudatus) belong to the genus Garra.