Acanthopagrus taiwanensis, Yukio Iwatsuki & Kent E. Carpenter, 2006

Yukio Iwatsuki & Kent E. Carpenter, 2006, Acanthopagrus taiwanensis, a new sparid fish (Perciformes), with comparisons to Acanthopagrus berda (Forsskål, 1775) and other nominal species of Acanthopagrus., Zootaxa 1202, pp. 1-19 : 4-15

Acanthopagrus taiwanensis n. sp.

New English Name: Taiwan Picnic Seabream

Figs. 1A, 2D -E, 3A -4A and 5A -B, Table 1

Holotype: MUFS 22854 , male, 167 mm SL, estuary basin of Tung-kang River (purchased in Tung-kang Fish Market), southwestern Taiwan , Y. Iwatsuki, hook-and-line (according to sellers in market), 22 May, 2005 .

Paratypes: (5 specimens) MUFS 11870 , sex not determined, 110 mm SL, Tung-kang, southwestern Taiwan , M. Akazaki, 25 February 1973 ; MUFS 22165 , sex not determined, 184 mm SL, Tung-kang, southwestern Taiwan , Y. Iwatsuki, 27 December 2002 ; MUFS 22166 , female, 216 mm, same data as MUFS 22165 ; MUFS 22855 , female, 175 mm SL, data same as holotype ; MUFS 22857 , sex not determined, 106 mm SL, mouth of Tungkang River, southwestern Taiwan , Y. Iwatsuki, shrimp set nets, 22 May, 2005 .

Diagnosis

Dorsal fin with 11 or 12 spines and 10 to 12 soft rays; anal fin with 3 spines and 8 or 9 soft rays; 3 ½ scale rows between the fifth dorsal-fin spine base and lateral line; 3 or 4 cheek scale rows; body rounded; ventral infraorbital series above posterior part of upper jaw changing from almost straight to weakly concave with growth (Figs. 2B, 3A); second anal-fin spine robust and dull pointed (Fig. 2C); upper and lower jaws with thin lips, especially anteriorly; broadly rounded rostral-most head squamation with around 10 somewhat smaller scales anteriorly (Fig. 4A); prominent ridge developing just before eye with growth, dorsal head profile becoming slightly convex from snout to just above eye with growth (Figs. 1A, 2B and 3A); upper and lower molar teeth strongly developed and flattened on both sides, gradually more molariform posteriorly (Fig. 5A -B); upper molars in 3 or 4 rows anteriorly and 4 or 5 posteriorly, rows generally irregular, outer third row markedly larger; lower molars in 3 or 4 rows anteriorly and 2 or 3 rows posteriorly, rows generally irregular, the innermost row largest posteriorly (outer second row gently curved outward)(Fig. 5C -D); head and body mostly black, sharply demarcated from whitish belly and chin (Figs. 1A and 3A); a dark spot at top of pectoral-fin base (Figs. 1A and 3A).

Description

Counts and measurements of the holotype and 5 paratypes are given in Table 1. In the following description, data for the holotype are presented first, followed by the 5 paratype specimens in parentheses where different. Characters stated in the diagnosis are not repeated.

Body compressed. Mouth somewhat oblique. Maxillary reaching to below middle of pupil and larger than eye diameter. Upper jaw protruding slightly in front of lower jaw. Teeth in jaws in 3 to 5 crowded rows, about 6 (6 or 7) curved canines anteriorly in the upper jaw and 6 in the lower jaw. Suborbital depth greater than dermal eye opening. Anterodorsal profile from above eye gently curved. Profile of occipital ridge distinct above eye. Dorsal-fin spines strong, length of first slightly more than half length of second; second slightly shorter than third; fourth or fifth spine longest. Longest soft dorsal-fin ray shorter than longest dorsal-fin spine. First anal-fin spine short, much less than orbit diameter; second spine robust, its length clearly less than length of head without snout and slightly longer than snout; third anal-fin spine shorter than second spine. First anal-fin ray somewhat shorter than second and third anal-fin spines. Pectoral-fin tip not reaching to vertical through first anal-fin spine base, its length clearly greater than head length. Pelvic fin with first ray somewhat produced, its length shorter than head; pelvic-fin spine longer than snout.

Color (from fresh specimens)

Head and body black (Fig. 1A). Both chin and belly markedly whitish in large specimens, especially over about 180 mm SL. Dorsal, caudal, anal, and pectoral fins black; ventral fins whitish. Color does not change much with preservation.

Distribution

Acanthopagrus taiwanensis is currently known only from Tung-kang, southwestern Taiwan. According to local fishermen and buyers, the species is common around southwestern Taiwan. As the first author did not observe the species in extensive field sampling and examination of museum specimens from Korea, Japan, China (including around Hong Kong and Hainan Island), Vietnam, and elsewhere around Southeast Asia, it may be endemic to Taiwan.

Etymology

The species’ name, “taiwanensis” reflects the type locality, Taiwan.

Discussion

Acanthopagrus taiwanensis and A. berda : Acanthopagrus taiwanensis and A. berda appear to be closely related and are perhaps sister species based on shared specializations. A cladistic analysis of species within this genus is premature as confusion remains about the status of some species and the relation of this genus to Sparidentex (Orrell & Carpenter, 2004). The combination of body shape, fin coloration and typical number of scale rows above the lateral line is unique for these species within Acanthopagrus , supporting a close relationship. In addition, the shape of the ventral edge of the first 2 infraorbitals is concave in larger individuals of both species, while it is straight in all other described species of Acanthopagrus . The concavity is much more pronounced in A. berda than A. taiwanensis and represents a clear difference between the 2 species.

Synonyms of Acanthopagrus berda : Acanthopagrus berda ( Forsskal, 1775) was described on the basis of a dried skin of the holotype (see Table 1; Fig. 2A) from Luhaiya, Yemen, Red Sea. Despite the condition of the holotype, the 4 most important diagnostic characters of this species are evident: a strong concavity of ventral edge of first 2 infraorbitals above posterior part of upper jaw (Figs. 1B, 2A and 3B); 3 ½ scale rows between fifth spinous dorsal-fin spine base and lateral line; anal, ventral, and caudal fins uniformly blackish; and second anal-fin spine longer than third spine. We have examined numerous specimens from throughout the Indo-Pacific, including the Red Sea, that correspond to these key characters. The concavity of the ventral edge of the first 2 infraorbitals is observed in specimens larger than around 130-150 mm SL. The holotype of Sparus hasta Bloch and Schneider (1801) has the same diagnostic characters as A. berda . Bauchot and Skelton (1986) determined this species to be a junior synonym of A. berda and we concur.

Sparus calamara Cuvier, 1829 (based on a drawing of Russell, 1803: pl. 92) is clearly a junior synonym of A. berda , with the following diagnostic characters evident on the drawing: a relatively deep body (52 % of SL, see Table 1); 3 ½ scale rows between the fifth spinous dorsal-fin spine base and lateral line; ventral edge of the first 2 infraorbitals above the posterior part of the upper jaw weakly concave; and black ventral, anal, and caudal fins (Table 1, Figs. 1i -5). Chrysophrys calamara Valenciennes in Cuvier & Valenciennes, 1830 (preoccupied by Cuvier, 1829), based on 3 syntypes has long been identified as a junior synonym of A. berda by subsequent researchers (Bauchot & Daget, 1972; Bauchot & Skelton, 1986). Day (1875) also described “ Chrysophrys berda var. calamara ” on the basis of a ZSI F1785’s specimen. However, since the name “ calamara ”, is preoccupied by Cuvier, 1829, Day’s “ calamara ” has no taxonomic significance in the context of A. berda .

Castelnau (1861) described Pagrus caffer (type probably lost, P. Pruvost, pers. comm.) from Durban (Port Natal), South Africa and Smith & Smith (1986) later considered it to be a synonym of A. berda ; we tentatively concur. We examined the holotype of Gilchrist & Thompson’s (1908) Chrysophrys robinsoni and its characters are consistent with A. berda , including the broadly rounded rostral-most head squamation with small scales anteriorly (as in Fig. 4B).

Comments on other nominal species assigned to Acanthopagrus : No type specimen is known for Coius datnia , described by Hamilton (1822) from the Ganges River mouth, India, but a fine plate (fig. 29 of pl. 9) accompanied the description. This species has long been considered a junior synonym of A. latus (Houttuyn, 1782) because of its yellow pelvic and anal fins (Kottelat, 1986, 2000). However, the absence of molar teeth given in the original description and plate strongly support its placement in the genus Sparidentex (Bauchot & Smith, 1983) and clearly distinguish it from A. berda and A. taiwanensis . Hamilton’s species needs further study.

Gilchrist & Thompson (1908) described Chrysophrys estuarius from South Africa. The three syntypes have a relatively slender body and a straight ventral edge of the first 2 infraorbitals above the posterior upper jaw, characters that distinguish it from A. berda and A. taiwanensis .

Chrysophrys schlegelii Bleeker, 1854 and Mylio butcheri Munro, 1949 are valid species in the genus Acanthopagrus from the East Asian Shelf and southern Australia, respectively (Akazaki, 1984; Allen et al., 2002; Cadwallader and Backhouse, 1983; Gomon et al., 1994). They are clearly distinct from both A. berda and A. taiwanensis with slender bodies and 4 ½-5 ½ scale rows between the fifth dorsal-fin spine base and lateral line (vs. 3 ½ for A. berda and A. taiwanensis ). Akazaki (1962) considered Acanthopagrus swinhonis czerskii Berg, 1914 as a junior synonym of A. schlegelii . We did not see the type specimens but followed his idea based on the detailed figure included in the description.

Chrysophrys australis Guenther, 1859 , C. swinhonis Guenther, 1874 , C. rubroptera Tirant, 1883 , Petrus belayewi Hora & Misra, 1943 , Acanthopagrus sivicolus Akazaki, 1962 , and Chrysophrys novaecaledoniae Castelnau, 1873 show 4 ½ scale rows between the fifth dorsal-fin spine base and lateral line (rarely 3 ½ in A. australis and yellowish pelvic- and anal-fin rays, as described in Carpenter, 2001, and 2 of 4 syntypes of C. swinhonis with 6 ½, as discussed below) on the basis of their type specimens. But C. rubroptera and A. sivicolus were confirmed by a fine photograph (Kottelat, 1986) and nontype specimens examined, respectively. In addition, these species have a straight ventral edge of the first 2 infraorbitals above the posterior part of the upper jaw, in contrast to strongly or weakly concave in A. berda and A. taiwanensis .

The taxonomic status of C. swinhonis has not been examined in detail since the original description by Günther (1859). In 4 syntypes (87-279 mm SL), the pored lateralline scales range from 51 to 56 (vs. 42 to 44 in A. berda and 43 or 44 in A. taiwanensis ), and scale rows between the fifth dorsal-fin spine base and lateral line are 4 ½ or 6 ½. Although a detailed review of C. swinhonis is still required, it is clearly distinct from both A. berda and A. taiwanensis .

The holotype of Roughleyia palmaris Whitley, 1935 has 4 scale rows between the fifth dorsal-fin spine base and lateral line, and the upper head profile gibbous (not gibbous in A. berda and A. taiwanensis ). Hutchins (2001) and Allen et al. (2002) regard it as a valid species of Acanthopagrus confined to southwestern Australia.

Although no types of Pagrus macrocephalus Basilewsky, 1855 are known at ZIN was considered to be a junior synonym of A. schlegelii (Bleeker, 1854) by Akazaki (1962) and Sadovy & Cornish (2000). However, Basilewsky’s (1855), figure 3, tab. I (referred to erroneously as tab. III, fig. 1, which is a sciaenid) is a sparid that is clearly a species of Pagrus , recognizable by its scaly interorbital area, red body, and 7 or 8 scale rows between the fifth dorsal-fin spine base and lateral line (Tables 1-2)(Akazaki, 1962). Pagrus macrocephalus needs further examination and may be a synonym of Pagrus major (Temminck & Schlegel, 1844) .

Richardson’s (1846) Chrysophrys xanthopoda and C. auripes from Canton, Chinese Seas are based on BMNH stuffed syntypes that have not been located (J. Maclaine, pers. comm.). However, detailed unpublished color plates of both species in the BMNH library have yellowish pelvic, anal, and caudal fins, especially along with lower margins. These nominal species may, therefore, be junior synonyms of A. latus , which has the same fin coloration (see description of Akazaki, 1962, 1984 and Carpenter, 2001).

Chrysophrys longispinnis Valenciennes in Cuvier & Valenciennes, 1830 , considered by Bauchot & Daget (1972) to be a junior synonym of A. berda , is unlikely to be that species. Two of 3 syntypes of C. longispinnis from Bengal have an extremely long second anal-fin spine (1.59-1.60 ratio of second anal-fin spine to third anal-fin spine vs. 1.23-1.50, mean 1.36 in A. berda , and 1.32-1.40, mean 1.37 in A. taiwanensis ; see Table 1). The third syntype from Japan has a relatively deep body and fine minute posteroventral serration of the preopercular flange characteristic of Japanese A. latus (Houttuyn, 1782) (Akazaki, 1962). The type series of C. longispinnis , therefore, includes 2 species, neither of which is A. berda or A. taiwanensis .

Chrysophrys cuvieri , described (Day, 1875) from Mangalore, India, has been synonymized with A. berda (Whitehead & Talwar, 1976; Randall, 1995; Ferraris et al, 2000), although the type designation is unclear. Whitehead & Talwar (1976) identified the following potential Day specimens of C. cuvieri : AMS B.8225, 129 mm SL, Madras; BMNH 1975.9.30.21, 109 mm SL, Cochin; RMNH 1060, not seen by us. The original description and type locality of C. cuvieri was apparently based on ZSI 1782 although the type was not seen by us. Bauchot & Smith (1983) synonymized C. cuvieri with Sparidentex hasta (Valenciennes in Cuvier & Valenciennes, 1830) using dentition inconsistent with Acanthopagrus as a key character. Day’s original description of C. cuvieri supports this contention since species of Sparidentex lack the large molariform teeth typically found in Acanthopagrus . Day’s (1875) description states: “Teeth -four to six sharp, pointed and rather conical incisors in front of either jaw, with villiform teeth behind them: a pointed and compressed row along the outer side of either jaw, the last few of which are small and with rounded crowns; internal to these are two rounded small molars in the lower and three in the upper jaw.” One of Day’s specimens of AMS B.8225 is clearly an example of Sparidentex without the typical molariform teeth found in Acanthopagrus . Although its status as a synonym of S. hasta needs to be resolved, C. cuvieri is clearly distinct from A. berda and A. taiwanensis .

Akazaki (1962) synonymized Sparus chrysopterus Kishinouye, 1907 (type locality: Kyushu, Shikoku, Inland Sea, Pacific coast of Hondo, Japan) with A. latus . Although the type appears to be lost at ZUMT, Kishinouye (1907) noted that the pelvic and anal fins of S. chrysopterus are yellow, a character typical of A. latus . The description of the molariform teeth of this species is similar to those of A. schlegelii and clearly places it in the genus Acanthopagrus , although distinct from A. berda and A. taiwanensis .