Centropomus parallelus Poey, 1860

Figueiredo-Filho, Jessé Miranda De, Marceniuk, Alexandre P., Feijó, Anderson, Siccha-Ramirez, Raquel, Ribeiro, Giovana S., Oliveira, Claudio & Rosa, Ricardo S., 2021, Taxonomy of Centropomus Lacépède, 1802 (Perciformes: Centropomidae), with focus on the Atlantic species of the genus, Zootaxa 4942 (3), pp. 301-338 : 321-323

Centropomus parallelus Poey, 1860

( Figure 13 )

Centropomus parallelus Poey, 1860: 120 (original description; Cuba)

Centropomus mexicanus Bocourt, 1868: 90 (original description; Oaxaca, Mexico)

Centropomus constantinus Jordan & Starks in Jordan & Evermann, 1896: 1125

Centropomus heringi Fowler, 1906: 425

Centropomus gabbi Fowler, 1906: 423

Centropomus pellegrini Puyo, 1936: 213

Type. MCZ 10307, 228 mm SL ( Fig. 13 ), loc.: Havana, Cuba, ex coll.: Poey.

Poey (1860) described C. parallelus based on 150 mm TL specimen, collected off Cienfuegos, Cuba. Howell-Rivero (1938) proposed that a specimen deposited in the Museum of Comparative Zoology (MCZ 10271) was the holotype. Rivas (1986) rejected the specimen as holotype, due to discrepancies in characters of the original description. The neotype (MCZ 10307), designated by Rivas (1986), is a 228 mm SL specimen also from Cuba, which conforms to the original description of the species.

Diagnosis. Centropomus parallelus differs from its congeners as follows:

Western Atlantic: from C. undecimalis , by having 60–88 (usually 64–70) scales on lateral line to caudal-fin base (vs 61–79, usually 66–75), 10–13 lower limb gill rakers on first arch (vs 8–11), 13–18 gill rakers on first arch (vs 11–14), the longest anal-fin spine usually reaching caudal-fin base when adpressed (vs longest anal-fin spine not reaching caudal-fin base), pelvic fin usually reaching the anus (vs pelvic fin not reaching anus), and a less elongate and higher body, with a mean body height of 26.6% SL (vs more elongate and lower body, with a mean body height of 22.9% SL; from C. irae , by having 60–88 (usually 64–70) scales on lateral line to caudal-fin base (vs 79–96), 10–13 lower limb gill rakers on first arch (vs 7–8), 13–18 gill rakers on first arch (vs 11–13), the longest anal-fin spine usually reaching caudal-fin base when adpressed (vs longest anal-fin spine not reaching caudal-fin base), pelvic fin usually reaching the anus (vs pelvic fin not reaching the anus), and a less elongate and higher body, with a mean body height of 26.6% SL (vs more elongate and lower body, with a mean body height of 22.3% SL); from C. ensiferus , by having 60–88 (usually 64–70) scales on lateral line to caudal-fin base (vs 48–58), 10–13 lower limb gill rakers on first arch (vs 10–15), 13–18 gill rakers on first arch (vs 15–20), and pelvic fin usually reaching the anus (vs pelvic fin usually not reaching the anus); from C. pectinatus , by having 60–88 (usually 64–70) scales on lateral line to caudal-fin base (vs 47–52), 10–13 lower limb gill rakers on first arch (vs 14–16), 13–18 gill rakers on first arch (vs 20–23), and anal fin with III,7 (vs III,8); from C. poeyi , by having 60–88 (usually 64–70) scales on lateral line to caudal-fin base (vs 69–74), 10–13 lower limb gill rakers on first arch (vs 9–12), pelvic fin usually reaching the anus (vs pelvic fin not reaching anus), and second dorsal fin with I,11 (rarely I,10) (vs I,9).

Eastern Pacific: from C. nigrescens , by having 60–88 (usually 64–70) scales on lateral line to caudal-fin base (vs 68–76), 10–13 lower limb gill rakers on first arch (vs 8–9), and the longest anal-fin spine usually reaching caudal-fin base when adpressed (vs longest anal-fin spine not reaching caudal-fin base); from C.viridis , by having 60–88 (usually 64–70) scales on lateral line to caudal-fin base (vs 67–75), the longest anal-fin spine usually reaching caudal-fin base when adpressed (vs longest anal-fin spine not reaching caudal-fin base), and second dorsal fin with I,11 (rarely I,10) (vs I,9); from C. armatus , by having 60–88 (usually 64–70) scales on lateral line to caudal-fin base (vs 47–59); from C. robalito , by having 60–88 (usually 64–70) scales on lateral line to caudal-fin base (vs 47–59), 10–13 lower limb gill rakers on first arch (vs 14–17), and 13–18 gill rakers on first arch (vs 21–25); from C. unionensis , by having 60–88 (usually 64–70) scales on lateral line to caudal-fin base (vs 46–52), 13–18 gill rakers on first arch (vs 15–20), the longest anal-fin spine usually reaching caudal-fin base when adpressed (vs longest anal-fin spine not reaching caudal-fin base), and second dorsal fin with I,11 (rarely I,10) (vs I,9); from C. medius , by having 60–88 (usually 64–70) scales on lateral line to caudal-fin base (vs 53–65), 10–13 lower limb gill rakers on first arch (vs 14–16), 13–18 gill rakers on first arch (vs 18–23), the longest anal-fin spine usually reaching caudal-fin base when adpressed (vs longest anal-fin spine not reaching caudal-fin base), and anal fin with III,7 (vs III,8).

Molecular data. Specimens of C. undecimalis forms a distinct cluster in the COI sequences analysis, with 100% bootstrap value ( Figure 8 ). The genetic distance between this species and other species of Centropomus ranged from 0.142 to 0.184 ( Table 7). The C. paralleus clade is the sister group of a large clade that includes C. undecimalis , C. irae , C. nigrescens , C. viridis , C. pectinatus and C. medius .

Description. Morphometrics and meristics data are summarized in Tables 8, 9 and 10 on the basis of 182 specimens, 44.4–432 mm TL (201.8 mm average). Body high, mean body height 26.6% SL (10.0%–31.0%); body slightly elongate, head moderately elongate, mean head length 37.6% SL (29.6%–71.3%); head moderately high, mean head height 24.2% SL (18.9%–28.4%); largest specimen measured 357 mm SL. Lower jaw protractile, slightly longer than upper jaw, extending beyond tip of snout. Snout narrow and long, mean snout length 11.4% SL (7.6%–22.5%); eye relatively small, mean orbital diameter 19.1% HL. Preoperculum heavily serrated in adults; operculum smooth, without spines, bordered by a membrane that extends beyond origin of pectoral fin; 3–8 small spines in dorsal part of operculum.

Dorsal fins separated by a small gap, not connected by membrane, first dorsal fin VII–IX, second dorsal fin I,10–I,12. Pectoral fin with 13–16 rays, longest ray extending beyond origin of pelvic fin. Pelvic fin I,5, longest ray usually reaching anus. Anal fin III,7 (rarely III,6 or III,8), second spine usually longer than first spine; longest spine usually reaching caudal-fin base when adpressed. Caudal fin deeply forked. Lateral line slightly arched with ctenoid scales, extending onto caudal fin. Lateral line with 60–88 scales to caudal-fin base, usually 64–70; 64–90 scales extending onto end of caudal fin. Scale rows between lateral line and origin of second dorsal fin 8–14; scale rows between lateral line and origin of anal fin 10–15. 8–13 (usually 10–13) lower limb gill rakers on first arch; 12–18 (usually 13–18) gill rakers on first arch, excluding rudiments; 12–21 (usually 18–20) gill rakers on second arch, including rudiments.

Distribution. Centropomus parallelus occurs between the southeastern portion of the Gulf of Mexico, the Antilles, and Santa Catarina, Brazil. The distribution of C. parallelus is in accordance with previous records in the literature ( Meek & Hildebrand 1925; Cervigón 1966; Figueiredo & Menezes 1980; Guitart 1985; Rivas 1986; Carvalho-Filho 1999).

Remarks. Rivas (1986) stated that it is possible to differentiate C. parallelus from C. mexicanus based on scale counts in lateral line, above lateral line, below lateral line, and around the caudal peduncle, and that both species are sympatric in the Atlantic. However, as shown here, these characters exhibit high overlap and large variation within species and thus are non-informative for species diagnosis. Rivas (1986) also stated that the number of scales in the lateral line is the main character for separation between the two species, with C. parallelus having 79–92 scales and C. mexicanus 68–78 scales. This character is also presented in the description of their type specimens, C. parallelus with 85 scales and C. mexicanus with 67 scales, although Rivas (1986) noticed counting discrepancies between the original description of C. parallelus and its holotype (MCZ 10271). Previous studies indeed indicate a strong variation in the number of lateral line scales of C. parallelus ( Meek & Hildebrand 1925 [77–90]; Cervigón 1966 [80–86]; Figueiredo & Menezes 1980 [70–75]; Guitart 1985 [79–89]; Rivas 1986 [79–92]; Carvalho-Filho 1999 [79–92]); the number of lateral line scales of C. mexicanus , in turn, is only mentioned by Carvalho-Filho 1999 [68–79] and Rivas 1986 [68–78]. Specimens of C. parallelus and C. mexicanus are morphologically very similar and have largely overlapping geographic distributions. The character variation found here in previously identified specimens of both species are: lateral line scales (66–88 in C. parallelus vs 64–86 in C. mexicanus ); scales above lateral line (9–12 in C. parallelus vs 10–11 in C. mexicanus ); scales below lateral line (11–14 in C. parallelus vs 12–15 in C. mexicanus ); lower limb gill rakers on first arch (10–13 in C. parallelus vs 11–12 in C. mexicanus ); gill rakers on first arch (13–18 in C. parallelus vs 15–17 in C. mexicanus ). It is clear that formely proposed diagnostic characters for the genus are highly overlapping between C. parallelus and C. mexicanus in this study and those presented by previous authors. Therefore, based on the molecular and morphological evidence retrieved here, C. mexicanus is herein proposed as a junior synonym of C. parallelus .

Figures

FIGURE 8. Best maximum likelihood tree of Centropomus species elaborated with partial COI sequences. Numbers in the branches represent the percentage of occurrence of each group in 1000 pseudoreplicates by bootstrap. Values lower than 50% are not showed. Log likelihood = -3507.12.

FIGURE 13. Neotype of Centropomus parallelus Poey, 1860, MCZ 10307, 228 mm SL. Photo: Meaghan Sorce, MCZ.