Crenicichla lacustris (Castelnau)

Kullander, Sven O. & Santos de Lucena, Carlos A., 2006, A review of the species of Crenicichla (Teleostei: Cichlidae) from the Atlantic coastal rivers of southeastern Brazil from Bahia to Rio Grande do Sul States, with descriptions of three new species, Neotropical Ichthyology 4 (2), pp. 127-146 : 131-135

Crenicichla lacustris (Castelnau)

Figs. 3-4

Cycla lacustris Castelnau, 1855: 19 , pl. 8 fig. 3 (Type-locality: Dique, ou étang près de Bahia).

Crenicichla dorsocellata Haseman 1911: 355 , pl. 63 (Typelocality: Campos, R. Parahyba).

Crenicichla biocellata von Ihering, 1914: 333 (Type-locality: Rio Doce, Espírito Santo).

Material examined. 164 specimens, 54.2–290 mm, all from Brazil. Bahia: Syntype of C. lacustris, MNHN A 9501 (1 of 2, 126.0 mm), dique, ou étang près de Bahia. USNM 318218 (2, 75.0– 171 mm), rio Buranhem, approx. 5–6 km SSE of Eunápolis, Eunápolis, ca. 16°25’S 39°35’W. Minas Gerais: rio Doce drainage (all the rio Doce lakes are between 19°35’– 19°55’S, 42°25’ 42°40’W. Lagoa Dom Helvécio is the largest, at 19°46’S 42°37’W, near the left bank of the upper rio Doce, slightly east of the city of Marlieria): Lagoa Dom Helvécio: DBAV 0072 (1, 172.6 mm); DBAV 0073 (1, 121.7 mm); DBAV 0426 (7, 54.2–220.0 mm); DBAV 0548 (2, 210.0–235.0 mm); DBAV 1147 (2, 112.0– 130.2 mm); DBAV 1148 (2, 125.0– 138.6 mm); DBAV 1149 (3, 99.3–206.0 mm); DBAV 1150 (2, 197.0– 206.3 mm); DBAV 1151 (2, 198.0– 233.2 mm); DBAV 1152 (1, 117.2 mm); DBAV 1153 (1, 141.5 mm); DBAV 1157 (1, 126.0 mm); DBAV 1158 (4, 118.4– 133.7 mm); DBAV 1159 (3, 101.5– 133.9 mm); DBAV 1161 (4, 116.0–163.0 mm); DBAV 1162 (2, 127.8– 163.6 mm); DBAV 1163 (2, 127.6– 138.4 mm); DBAV 1164 (2, 139.8– 144.8 mm); DBAV 1165 (1, 162.0 mm); DBAV 1166 (7, 111.5– 119.7 mm); and DBAV 1167 (2, 115.5– 124.7 mm); DBAV 1168 (3, 102.4–135.0 mm). Lagoa Carioca: DBAV 0637 (1, 114.4 mm); DBAV 0642 (2, 87.7–119.7 mm); DBAV 1154 (1, 119.2 mm); DBAV 1155 (3, 87.5–122.8 mm); DBAV 1156 (7, 96.6–115.0 mm); DBAV 1160 (1, 122.9 mm); DBAV 1169 (3, 119.0– 132.8 mm); DBAV 1170 (3, 56.7–86.7 mm); DBAV 1171 (2, 80.0– 94.7mm); and DBAV 1172 (2, 96.4–98.0 mm). Rio Paraíba do Sul drainage: MNRJ 14390 (4, 84.8–114.6 mm), rio Novo, 6 km above of the confluence with rio Pomba, approx. 21°24’S 42°46’W; UFRJ 3731 (1, 144.8 mm), rio Paraíba do Sul,Além Paraíba, 21°53’S 42°40’30"W. Espírito Santo: MCP 18242 (4, 59.3–89.0 mm), rio Itabapoana, Bom Jesus do Itabapoana; MCP 18143 (2, 21.9–74.7 mm), rio São José das Torres on the road BR 101, between São José das Torres and Travessão, 21°04’43"S 41°14’02"W; MCP 18144 (6, 14.4–98.4 mm), rio Novo do Sul (tributary of rio Moa) on the road BR-101 just to the south of Rio Novo, rio Itapemirim drainage, 20°52’33"S 40°57’50"W; MNRJ 5274 (2, 78.0– 105.3 mm), lagoa Juparanã, rio Doce drainage, road Linhares-São Mateus, mun. Linhares, ca. 19°19’S 40°05’W; holotype of C. biocellata, MZUSP 1167 (1, 215.0 mm total length), lower rio Doce; paratype of C. biocellata, MZUSP 1781 (1, 150.0 mm), rio Santa Maria da Vitória, Santa Leopoldina, ca 20°06’30" 40°32’W; MZUSP 27570 (1, 96.8 mm), rio São José das Torres, road BR-101 Campos-Cachoeiro do Itapemirim. Rio de Janeiro: DBAV 095 (1, 108.0 mm), rio Paraíba do Sul drainage, near Resende, ca. 22°28’S 44°27’W; DBAV 096 (1, 101.0 mm), same locality; DBAV 0101 (1, 76.7 mm), represa do Funil, ca. 22°33’S 44°35’W; DBAV 0247 (2, 71.0– 124.6 mm), near Resende, ca. 22°28’S 44°27’W; DBAV 0854 (1, 290.0 mm), rio São João, Silva Jardim; DBAV 1173 (2, 116.9– 134.3 mm), rio Pirapitinga, Resende, 22°40’S 42°23’W; holotype of Crenicichla dorsocellata, FMNH 54155 (1, 177.0 mm) (previously CM 2721), rio Paraíba do Sul, Campos, ca. 21°46’S 41°19’20"W; MNRJ 13394 (11, 61.1– 149.9 mm), lagoa Brejo Grande, Campos; MNRJ 14391 (1, 143.3 mm), lagoa Feia, Macaé, ca. 22°S 41°22W; MNRJ 14392 (1, 124.8 mm), stream on road BR-393, 10 km from Monte Alegre, mun. Santo Antônio de Pádua; MNRJ 14393 (1, 157.5 mm), valão do rio São Domingos, at São José de Uloa, 21°21’S 41°57’W, Itaperuna, rio Muriaé system; MNRJ 14394 (6, 91.3–173.7 mm), ribeirão São Domingos, at São José de Uloa, 21°21’S 41°57’W, rio Muriaé system, Itaperuna; MNRJ 14395 (8, 4 measured, 162.7–222.4 mm), represa do Funil, Resende, ca 22°33’S 44°35’W; MNRJ 14396 (7, 18.9–37.3 mm), rio Carangola, left bank tributary of rio Muriaé at Bridge on road BR 356, Itaperuna, ca. 21°12’S 41°55’W; MNRJ 14397 (1, 125.7 mm), córrego Romão, Romão, km 2 on road BR-393, Cambuci, ca. 21°33’30"S 41°55’W; MNRJ 14398 (1, 123.6 mm), rio Pirapitinga, Resende, 3 or 4 km from the mouth, Monte Alegre; MNRJ 14399 (1, 190.1 mm), lagoa de Cima, ca. 21°46’28"S 41°31’40"W; MZUSP 3150 (1, 101.3 mm), rio Muriaé, Cardoso Moreira, 21°29’S 41°38’W; MZUSP 41203 (1, 72.3 mm), same data; MZUSP 41215 (2, 96.8–120.8 mm), rio Paraíba do Sul, São João da Barra, ca. 21°38’S 41°03’30"W; MZUSP 41216 (1, 117.8 mm), córrego Pedra D’água, São Fidélis, 21°39’20"S 41°45’40"W; UFRJ 3733 (2, 113.3–224.0 mm), UFRJ 3736 (1, 113.7 mm), and UFRJ 3737 (1, 158.0 mm), rio Muriaé, left bank tributary to rio Paraíba do Sul, near crossing of roads BR 356 and RJ-186, Itaperuna, 21°12’30"S 41°53’30"W; UFRJ 3740 (1, 111.9 mm), rio Grande, source of rio Dois Rios (right bank tributary of rio Paraíba do Sul), 2 km from bridge of road RJ-172, fazenda Humaita, São Sebastião do Alto, ca. 22°01’S 42°08’W; UFRJ 3744 (2, 90.0– 100.2 mm), rio Dois Rios, right bank tributary to rio Paraíba do Sul, fazenda Poço d’Antas, São Fidélis.

Diagnosis. An elongate, large sized species of the C. lacustris group, distinguished from all other coastal southeastern Brazilian species by its color pattern. Lateral band continuous from head to caudal fin base, vs. subdivided into a series of blotches in C. punctata and C. maculata ; suborbital stripe short and narrow, occasionally obsolete, vs. wide and prominent in C. iguapina ; small dark spots on side of head present, vs. absent in C. tingui ; in C. tingui both males and females have side of body spotted, though in males spots not extending onto abdominal side and lateral band remaining prominently pigmented, whereas in C. lacustris males lateral band indistinct in combination with dark spot pattern. From other species of the C. lacustris group, C. lacustris can be distinguinshed by E1 row scales count, 60-75 (vs. 47-65) and presence (vs. absent) of small dark spots on side of head present.

Description. Based primarily on specimens over 100.0 mm. Largest male 235.0 mm, largest female 290.0 mm. Measurements given in Table 1 , counts in Tables 4-8 . See Figs. 3-4 for general aspect.

Comparatively elongate, body depth 18.6-23.9% SL. Head about as deep as wide or slightly deeper than wide. Caudal peduncle longer than deep. Snout long, rounded when viewed from above, pointed in lateral view. Lower jaw prognathous. Ascending premaxillary process reaching to 1/2 of orbit. Maxilla reaching to vertical from anterior margin of orbit. Upper lip thick and wide, folds not continuous but cutting into symphyseal wide thickening. Postlabial skin fold margin truncate or slightly rounded. Orbit supralateral, not visible from below, chiefly in anterior half of head. Nostrils dorsolateral, about halfway between orbit and margin of postlabial skin fold and with low tubular margin but no anterior marginal skin flap. Preopercle regularly serrated.

Flank scales strongly ctenoid. All scales on head, anteriorly on back, along dorsal fin base, chest, and on belly below line from lower edge of pectoral axilla to anal fin origin and along anal fin base cycloid. Predorsal scales small, superficially embedded in skin, extending forward almost to transverse frontal lateralis canal. Prepelvic scales very small, superficially embedded in skin. Cheek fully scaled or narrowly naked ventrally and anteroventrally; 7-11 rows of scales below eye, embedded in skin. Interopercle naked. Circumpeduncular scale rows 11-15 dorsally, 12-16 ventrally (total 28- 33, including lateral lines).

Scales between upper lateral line and dorsal fin base 9-12 anteriorly, 4-6 posteriorly; 3-4 scale rows between lateral lines. Anterior upper lateral line scales slightly larger and more elongate than adjacent scales, remaining lateral line scales nearly same size as adjacent scales; 3 scales impinging on each scale of anterior part, 2 on each scale of posterior part of upper lateral line; 2 scales impinging on each scale of lower lateral line. Scales absent from dorsal, anal, pectoral, and pelvic fins. Caudal fin squamation concave, marginally extending beyond middle of fin.

First dorsal spine about 1/3-1/4 length of last; spines increasing in length to last but subequal from about 10 th. Soft part of dorsal fin pointed; both sexes sometimes with produced middle rays, reaching to about base of caudal fin. Soft anal fin with acuminate tip; not reaching to caudal fin base in both sexes. Caudal fin rounded. Pectoral fin rounded, reaching about halfway to anal fin origin. Pelvic fin inserted well posterior to vertical from pectoral axilla, with acuminate tip, second ray longest, reaching about halfway to anal fin origin; anterior rays and margin thickened.

All teeth pointed, slightly recurved. Outer row of teeth distinctly larger than inner teeth and larger anteriorly than posteriorly. Upper jaw with 4-5 inner rows; all teeth depressible. Lower jaw anteriorly with 3 inner rows; all teeth inclinable or depressible.

Microbranchiospines present externally on second through fourth gill arches.

Coloration in alcohol. Yellowish to brownish, darker on back, whitish on belly, with brown or black markings. Lateral line scales light. Upper lip gray. Dark brown preorbital stripe from orbit to margin of mouth. Dark brown postorbital stripe runnning straight from orbit to dorsal end of gill cleft, sometimes reduced to blackish spot immediately posterior to orbit. Nuchal markings very light in medium and large specimens. In specimens up to 116 mm (MNRJ 13394) sometimes with black blotch along posttemporal and dark band along margin of predorsal scales above sphenotic.

Suborbital stripe black, never extending below middle of cheek, and sometimes reduced to small spot at orbital margin. Occasionally ends in few dots scattered down cheek for three or four scales. Suborbital stripe better developed in females than in males. Males with body covered with seemingly irregularly scattered blackish spots, except for ventral part of abdomen. Such spots found also on opercle and subopercle and posteriorly on cheek. Small specimens (to ca. 75 mm), however, lack spots on head. Females have spots on body or completely lack body spots. Both sexes with wide horizontal dark stripe extending from gill opening to end of caudal peduncle, restricted to area between lateral lines, but generally more distinct in females. Females also generally show 5-7 well expressed vertical bars extending from lateral band to dorsal fin base, and another two dark blotches across dorsal margin of caudal peduncle.

Dorsal fin in females smoky with wide dark margin and sometimes with distinct light submarginal stripe. Most females with slightly elongate, ocellated black blotch with white margin on middle portion of dorsal fin, usually between 13 th and 17 th or 12 th and 17 th spines; one female (USNM 318218) show ocellus between 11 th and 14 th spines and another small one between 14 th and 17 th, in small females blotch sometimes indistinct and without clearly defined light margin. Soft portion of dorsal fin may feature few dark dots. Males with dorsal fin grayish with numerous dark spots irregularly distributed over both spinous and soft portions.

Anal fin grayish; males with numerous dark spots on posterior membranes; in females anal fin immaculate or with only few dark spots. Pelvic fins whitish. Females with only few dark spots on caudal fin. Caudal fin of males vividly patterned with dark dots. Rounded caudal spot black and ocellated. One male specimen (MNRJ 14394) with dark dots between pectoral fin rays, otherwise pectoral fin immaculate.

Color in life. Two large adults, male and female, ca. 220 and 230 mm, from lagoa de Juturnaiba, State of Rio de Janeiro, collected July 1981 and photographed freshly preserved, similar to preserved specimens. However, dark spots on side and head maroon, more reddish in male, and female with orange horizontal stripe along side of abdomen adjacent to lateral band. Lateral band conspicuous in both specimens. Young female from rio do Salto, Macaé, State of Rio de Janeiro, collected 28 January 1997 also with well marked lateral band and many large reddish spots on sides and gill cover.

Variation. Data is given separately here for the sample from the rio Doce lakes. These specimens are in a poor state of preservation and the only two specimens sexed are definitely females. These specimens differ from other C. lacustris in having a longer snout, a more prognathous (and longer) lower jaw, and a modal anal fin count of III,8 rather than III,9. None of the specimens features a lateral band, as is always evident in female C. lacustris from elsewhere and commonly also apparent at least as a trace in males. Instead both sexes have spots all over the body, the females also featuring dark spots on the gill cover. Although Menezes (1987) was able to distinguish the sympatric rio Doce Oligosarcus as a species distinct from the species of Oligosarcus in the Paraíba do Sul and coastal plain, we hesitate to give species status to the rio Doce Crenicichla for want of well preserved material that would enable us to provide a useful diagnosis. With the material at hand, the overlap of counts and meristics prevents us from a satisfactory diagnosis, and the state of preservation makes a proper evaluation of the color pattern equally unsatisfactory.

Stomach contents. Two stomachs examined, DBAV 0426 (154.2 mm) and MNRJ 14390 (84.8 mm), contain fish remains.

Geographical distribution. Collections range from coastal rivers from rio Buranhem south to rio Paraíba do Sul, including the rio São João drainage in the State of Rio de Janeiro ( Fig. 2 ). There is one population in the lakes of the upper rio Doce, but otherwise no material is available from the upper rio Doce valley. The species apparently occurs all along the rio Paraíba do Sul downstream from the area of Resende. The species is absent from the rio Mucuri, which is within the total range of C. lacustris .

Notes. The two syntypes of C. lacustris were reported by Castelnau to come from Bahia (= Salvador, State of Bahia). As no Crenicichla lacustris group species has been found again further north than the rio Buranhem, we find it likely that the locality given by Castelnau is in error. Castelnau lived in Salvador as the French Consul in 1849 and wrote the multivolume itinerary of his South American travels in that period ( Papavero, 1975). Circumstantial evidence suggests, however, that no C. lacustris group species occur in the Salvador area, and that the syntypes of C. lacustris most likely came from near Rio de Janeiro where Castelnau stayed from 17 June till 13 October 1843.

Recent extensive collecting efforts in the rio Jequitinhonha and rivers from the Jequitinhonha north to include the rio Paraguaçu (by J. Garavello, S. Jewett and R. Macedo, and R. E. Reis et al.) failed to locate any Crenicichla material. The Buranhem specimens (the next river south from the Jequitinhonha south) reported herein likely are the most northern Crenicichla treated here, and they conform to C. lacustris .

Castelnau made his most important collecting trip in South America in 1843, starting from Rio Janeiro, travelling along the Tocantins, then west to Peru and finally down the Amazon and through to Guyana from where he returned to France in 1847.

Castelnau (1855) reported mostly on marine fishes observed or collected at Salvador and Rio de Janeiro, and freshwater fishes mostly collected or observed in the Araguaia and in Peru, and in inland Minas Gerais. One species is explicitly reported from Rio de Janeiro, viz. Loricaria castanea (p. 46, currently in the genus Loricariichthys ). Erythrinus trahira (presumably an Hoplias species) is reported as common in the fresh waters around Salvador (p. 56), Chromys paraguassu and C. obscura , representing a single valid species of Geophagus , are described from the rio Paraguaçu, which has its mouth at Salvador. Chalceus devillei (p. 69; now in the genus Brycon ) and Tetragonopterus vittatus (p. 66) are described as new species from Bahia, and Xiphorhynchus hepseticus Müller & Troschel is reported (p. 75) from Bahia.

Howes (1982) redescribed Brycon devillei , emphasizing the distinctiveness of the species. Howes tentatively included material from the rio Jequitinhonha ( Brycon insignis Steindachner ) and rio Doce. Another species described from ‘Bahia’, viz. B. bahiensis Günther , is apparently known only from the type specimen and one other more specimen from Rio de Janeiro ( Howes, 1982). Lima (2003) assigned it as a synonym of Brycon opalinus (Cuvier) , a species with the type locality Rio de Janeiro. Howes (1982) grouped B. devillei with B. acuminatus , B. ferox , and B. reinhardti .

Brycon ferox Steindachner is known only from the type material collected in the rio Mucuri. Howes (1982) thought that it might be the same as Brycon acuminatus (Eigenmann & Norris) (synonym of B. insignis Steindachner in Lima, 2003), known only from two specimens, both collected in the rio Tietê, a tributary of the Paraná (but the type locality was erroneous, the holotype apparently being collected in the rio Paraíba do Sul, see Howes, 1982 and Lima, 2003).

According to Lima (2003) Brycon reinhardti Lütken occurs in the São Francisco (rio das Velhas) and is assigned as a synonym of B. nattereri Günther (Paraná river basin).

Tetragonopterus vittatus is an Astyanax bimaculatus -like species, treated as a subspecies of A. bimaculatus by Fowler (1948) and later a provisional synonym of that species by Lima et al. (2003). It was reported from many localities near Salvador by Eigenmann (1921).

Menezes (1987) included Castelnau’s Xiphorhynchus hepseticus in his Oligosarcus acutirostris , which was reported from many localities between the mouth of the rio Jequitinhonha south to the rio São José das Torres, State of Espírito Santo, where it is replaced by O. hepsetus , which ranges all the way to the rio Cubatão in Santa Catarina.

Whereas from the above there is little doubt that Castelnau had freshwater fish collections from Bahia (better understood as the present city of Salvador than as the “State of Bahia”), viz. specimens of Geophagus , Brycon devillei , and Astyanax vittatus , it is equally clear that he had freshwater fish collections from Rio de Janeiro ( Loricariichthys castaneus ).

We find it likely that the ‘ X. hepseticus’ did not come from Salvador since there are no Oligosarcus north of the Jequitinhonha. Indeed, given that Castelnau had collections from Rio de Janeiro, it may have come from there rather than from Bahia, and represent O. hepsetus rather than O. acutirostris . Menezes did not examine Castelnau’s specimens and may have relied entirely on the locality for the identification.

It is therefore possible that the Crenicichla lacustris came from somewhere near Rio de Janeiro. There is no indication or evidence known to us that Castelnau sampled along the coast between Rio de Janeiro and Bahia.

Crenicichla dorsocellata is based on the holotype from Campos and a paratype from Santarém, in the Amazon drainage. The latter has since been identified by most authors as a member of the C. wallacii group which is common in the Amazon basin, and also features a prominent dorsal fin ocellus. Crenicichla dorsocellata falls within the variation of C. lacustris as here understood, and is therefore synonymized with that species.

Crenicichla biocellata was described from two specimens, one from rio Doce, collected by E. Garbe in 1906 and expressly referred to as the type; the other specimen was from Porto Cachoeiro, collected by E. Garbe in 1912. The type material of C. biocellata was not listed by Miranda Ribeiro (1918) in his catalogue of fishes in the Museu Paulista (presently MZUSP). Britski (1969: 210) listed the paratype (MZUSP 1781), noting that Pôrto Cachoeiro had meanwhile changed its name to Santa Leopoldina and is situated on the rio Santa Maria da Vitória and not on the rio Doce as stated in the original description. Ploeg (1991) also listed this paratype, but as the holotype. Actually, the holotype was present in the MZUSP but went undetected until we were able to identify it with the help of O. Oyakawa as a specimen of 215 mm total length in a lot of two specimens (MZUSP 1167), with collecting data as the lower rio Doce, E. Garbe 1906. The specimen conforms to Ihering´s description and has been separated as the holotype of C. biocellata , with the catalogue number MZUSP 1167.

We have not found any clear evidence of a specific distinctness of coastal plain material representing the nominal taxon C. biocellata and C. lacustris from the rio Paraíba do Sul and therefore synonymize C. biocellata with C. lacustris .

Whereas in preserved specimens males and females differ in the extent of the spotting of the sides, it appears as if the reddish spots of females are lost, whereas in males they are retained in preservative probably because they are underlain by more persistent melanophores. These diagnostic spots are likely a part of the breeding color pattern, unlike the head spots found in C. punctata , which are persistent in preservative and relatively smaller. Inasmuch as the other coastal species lack corresponding spots, the red spots on the gill cover, and perhaps those on the anterior body sides, are probably autapomorphic for C. lacustris .

Figures

Fig. 3. Crenicichla lacustris, male, MNRJ 14394, 173.7 mm SL, ribeirão São Domingos, rio Muriaé drainage, State of Rio de Janeiro, Brazil.

Fig. 4. Crenicichla lacustris, female, USNM 318218, 171.0 mm SL, rio Buranhem, State of Bahia, Brazil.

Fig. 2. Collecting localities of Crenicichla lacustris and C. mucuryna in the coastal rivers of eastern Brazil. Mostimportant rivers are:1. Rio Paraíba do Sul; 2. RioDoce; and 3. Rio Jequitinhonha.A symbol may cover more than one collecting site.