Curimata cerasina Vari 1984

Inhabits rivers, ponds and other highly turbid still waters, usually over muddy bottoms (Ref. 34109).

Curimata cerasina Vari

Curimatus schomburgkii.—Mago-Leccia, 1967:254 [Venezuela: Llanos of Río Orinoco basin].

Curimata sp.—Roman, 1983:97 [Venezuela: Río Orinoco basin; common name].

Curimata cerasina Vari, 1984b:30, fig. 1 [type locality: Venezuela, San Fernando de Apure].—Machado-Allison, 1987:48, 68, 72, 93–95 [fecundity, breeding, development], Machado-Allison et al., in Machado-Allison, 1987:135 [common name].—Vari, 1989, tables 2, 3 [phylogenetic relationships].

DIAGNOSIS.—Curimata cerasina can be distinguished from other species of Curimata, and indeed all other curimatids by the very wide interorbital region which is 0.53–0.58 of HL in contrast to 0.36–0.52 for all other species of Curimata. Only Curimata aspera, a species endemic to the Rio Amazonas basin, has an interorbital width (0.46–0.52) approaching, although distinct from, that of C. cerasina (see Figure 38). Differences in gape width (0.27–0.35 in C. aspera vs. 0.35–0.40 for C. cerasina) and the more robust body of C. cerasina further separate these forms.

DESCRIPTION.—Body moderately elongate, robust, more so in larger specimens, particularly females in spawning condition. Dorsal profile of head very slightly concave above orbit in smaller specimens, straight in larger individuals. Dorsal profile of body convex from rear of head to origin of rayed dorsal fin; straight and posteroventrally slanted at base of dorsal fin, gently convex from base of last dorsal-fin ray to caudal peduncle, convexity more pronounced in larger specimens. Dorsal surface of body with indistinct median keel anterior to rayed dorsal fin, smoothly rounded transversely posterior to fin. Ventral profile of body gently curved from tip of lower jaw to caudal peduncle, convexity more pronounced posteriorly. Prepelvic region flattened, margined laterally by distinct, nearly right, angles in body wall that extend from level of vertical through origin of pectoral fin to origin of pelvic fin. Median series of enlarged scales in prepelvic region flanked on each side by series of enlarged scales that conform in shape to lateral angle of prepelvic portion of body. Well developed median keel posterior to pelvic fin insertion with secondary obtuse keel about two scales dorsal of ventral midline on each side of postpelvic portion of body.

Greatest body depth at origin of rayed dorsal fin, depth 0.41–0.48 [0.44]; snout tip to origin of rayed dorsal fin 0.52–0.56 [0.53]; snout tip to origin of anal fin 0.81–0.85 [0.85]; snout tip to origin of pelvic fin 0.52–0.58 [0.54]; snout tip to anus 0.78–0.82 [0.81]; origin of rayed dorsal fin to hypural joint 0.52–0.58 [0.58]. Rayed dorsal fin pointed, anteriormost rays 3.7–5.3 times length of ultimate ray. Pectoral fin pointed; length of pectoral fin 0.17–0.25 [0.24], extends to or slightly beyond vertical through origin of pelvic fin. Pelvic fin pointed, length of pelvic fin 0.17–0.25 [0.24], reaches three-quarters of distance to origin of anal fin. Caudal fin forked. Adipose fin well developed. Anal fin emarginate, anteriormost branched rays 2.5–3.2 times length of ultimate ray. Caudal peduncle depth 0.12–0.15 [0.14].

Head obtusely pointed in profile, very wide; head length 0.30–0.35 [0.32] in specimens over 75 mm SL (0.37–0.38 in specimens of 45.0–55.0 mm SL); upper jaw longer, mouth inferior, teeth present in larvae, but absent in adults (Machado-Allison, 1987:72), snout length 0.32–0.39 [0.36]; nostrils of each side of head very close, anterior circular, posterior crescent shaped with aperture largely closed by flap of skin that separates nares; orbital diameter 0.25–0.32 [0.28]; adipose eyelid present, most developed anteriorly, with vertically ovoid opening over middle of eye; length of postorbital portion of head 0.39–0.46 [0.44]; gape width 0.34–0.39 [0.39], specimens under 18.0 mm SL with single series of conical teeth in each jaw, no teeth in jaws in larger specimens; interorbital region very wide, width 0.53–0.58 [0.57].

Pored lateral-line scales from supracleithrum to hypural joint 42 to 46 [45]; canals in scales straight; all scales of lateral line pored in larger specimens, individuals under 33.0 mm SL with lateral line incompletely pored, 4 to 6 series of scales extend beyond hypural joint onto caudal fin base; 13 to 15 [14] scales in transverse series from origin of rayed dorsal fin to lateral line, 8 to 11 [9] scales in transverse series from the lateral line to origin of anal fin; 12 to 15 [14] enlarged scales in midventral series anterior to insertion of pelvic fins; scales somewhat ctenoid, ctenii more developed in larger specimens, more so on ventral portion of body.

Dorsal-fin rays ii,9–10 (ii–10 rare), iii,9 [ii,9]; anal-fin rays ii,10–11, iii,10–11 [ii,11]; pectoral-fin rays 14 to 16 [15]; pelvic-fin rays i,8–9 [i,9].

Total vertebrae 31 (2), 32 (37), 33 (1).

COLOR IN LIFE.—Overall body coloration bright silver. A relatively large patch of intense red pigmentation on ventral and lateral surfaces of body surrounding region of origin of pelvic fin. Red coloration extends anteriorly to vertical through origin of pectoral fin, posteriorly two-thirds of distance from origin of pelvic fin to anus, slightly further in larger specimens, and dorsally to level of horizontal through origin of pectoral fin. Pelvic fins intense red. Dorsal fin rays edged with red. Caudal and anal fin rays red distally.

COLOR IN ALCOHOL.—Overall coloration in specimens that retain guanine on scales silvery or silvery-golden; darker on dorsal portions of head and body. Specimens that lack guanine on scales pale yellow to brown, darker on dorsal portions of head and body. Chromatophores more concentrated on scale centers in some specimens; darker pigmentation aligned to form faint longitudinal bands along scale series. No pronounced pigmentation pattern on head or body. Margins of median fins dusky, fin rays outlined by series of small chromatophores. Pelvics somewhat dusky in some individuals.

DISTRIBUTION.—Central portions of Río Orinoco drainage basin (Figure 40).

ECOLOGY.—Curimata cerasina inhabits rivers, ponds and other highly turbid still waters, typically over muddy bottoms. The species is common in the lower lying regions of the Llanos (savannahs) that are most subject to seasonal flooding (Machado-Allison, 1987:68; pers. observ.). Females with ripe ovaries were present in collections made in late March and April, with each female producing over 100,000 eggs (Machado-Allison, 1987:48, 68). The species breeds in the main channels of the watercourses in the Llanos, with juveniles of 17.0–60.0 mm SL captured in floating and emergent vegetation in late July and August. The larvae of the species feed on copepods and cladocerans utilizing oral dentition that is lost ontogenetically (Machado-Allison, 1987:72).

COMMON NAME.—Bocachica (Roman, 1983:97), Boca chico (Machado-Allison et al., in Machado-Allison, 1987:135).

MATERIAL EXAMINED.—96 specimens (87, 17.0–167.3 mm SL).

VENEZUELA. Apure: Flooded area along road from San Fernando de Apure to Arichuna, 15 km southeast of San Fernando de Apure, MBUCV V-14025, 1, (149.6, holotype of C. cerasina); USNM 257086, 3 (125.6–139.8, paratypes of C. cerasina); AMNH 54629, 1 (133.3, paratype of C. cerasina). Río Arauca, La Trinidad de Arauca, MBUCV V-5364, 4 (115.8–120.2). Río Apure Viejo at San Fernando de Apure, MBUCV V-8256, 1 (54.7). Inundated region off Río Apure Viejo, Hacienda La Guamote, near San Fernando de Apure, MBUCV V-11103, 6. Boca Apurito, Río Apure, MBUCV V-9120, 1 (146.0, paratype of C. cerasina). Río Apure, mouth of Río Apurito near San Fernando de Apure, MBUCV V-9015, 3 (113.5–167.3, paratypes of C. cerasina); ANSP 150195, 1 (145.8, paratype of C. cerasina). Canã el Pavoncito, Río Apure, MBUCV V-2779, 3 (101.7–111.4). Río Apure in front of airport at San Fernando de Apure, USNM 257087, 2 (151.3–160.9, paratypes of C. cerasina); AMNH 54628, 1 (145.1, paratype of C. cerasina). Río Apure, Los Boquerones, MBUCV V-635, 23 (105.1–135.8). Río Apure at junction with Río Portuguesa, MBUCV V-9334, 1 (111.7, paratype of C. cerasina). Canõ off Río Apurito, USNM 258258, 1. Río Arauca, USNM 257526, 2 (1 cleared and counterstained for cartilage and bone). Guarico: Río Tiznados, MBUCV V-2844, 1 (146.8). Laguna los Laureles between Camaguan and Canō Falcon, MBUCV V-6607, 7 (74.7–163.9, paratype of C. cerasina); USNM 257085, 1 (89.3, paratype of C. cerasina, cleared and counterstained for cartilage and bone); CAS 52169, 2 (90.2–97.4, paratypes of C. cerasina); FMNH 94591, 1 (87.9, paratype of C. cerasina). Canō Falcon, MBUCV V-6627, 10 (109.7–159.7). Esteros de Camaguan approximately 8 km north of Camaguan on highway from Calabozo to San Fernando de Apure, MBUCV V-5820, 1 (121.2); MBUCV V-11845, 1 (105.4, paratype of C. cerasina); MBUCV V-11893, 1 (146.0). Flooded borrow pit along side of road from Calabozo to Camaguan, USNM 257088, 2 (91.7–93.6, paratypes of C. cerasina); BMNH 1983.3.1:8, 1 (86.0, paratype of C. cerasina). Río Portuguesa near La Manga, Camaguan, MBUCV V-8691, 3 (134.0–148.6, paratypes of C. cerasina); BMNH 1983.3.1:7, 1 (131.7, paratype of C. cerasina). Río Portuguesa at Camaguan, USNM 257084, 2 (85.7–97.5). Río Portuguesa, Laguna Boca Ruido, MBUCV V-5807, 3 (17.0–45.6). Río Manapire, deep pool isolated from the main river channel, La Vega near Santa Rita, MBUCV V-5743, 2 (104.3–106.7, paratypes of C. cerasina); MBUCV V-5731, 1 (95.7). Portuguesa: Canō Maraca at bridge on road from Guanarito to Guanare MCZ 54344, 1, (88.3, paratype of C. cerasina). Barinas: Río Guanare Viejo, Boca de Tortumal, west of La Union, MBUCV V-3631, 1 (92.9).

Phylogenetic Biogeography

The species of the genus Curimata inhabit a significant portion of the total range of the family Curimatidae, which in turn occurs through much of South America. Such a broad geographic range in conjunction with a nearly resolved intrageneric phylogeny and detailed distributional information should provide insight into the historical biogeography of the genus, and to lesser degrees into those of the family and the lowland South American freshwater fish fauna.

Vari (1988), focusing on questions of areas of endemism of the Neotropical freshwater fish fauna and alternative hypotheses concerning the relative extent of speciation before and after the uplift of the Andes, discussed some aspects of the zoogeography of Curimata and the entire Curimatidae. Suprageneric topics will not, as a consequence, be detailed again in this study. Two primary questions can be addressed by the phylogenetic and distributional data at hand for Curimata. First, what are the historical associations of the river basins of South America inhabited by the species of Curimata as reflected in the hypothesized relationships of those taxa? Second, to what degree has secondary dispersal been a factor in the history of the members of the genus?

Figure 42 is an area cladogram for the species of Curimata in which the species names have been replaced by the areas that they inhabit, a system which simplifies the visualization of geographic distributions across the phylogenetic scheme. Although reduced area cladograms are typically used in such historical biogeographic studies I have chosen not to follow that practice since that would decrease the degree of available information relative to the second question—to what extent has dispersal been a factor in the distribution of the known species of Curimata?

Reference to Figure 42 shows that the historical biogeography of the genus has apparently been quite complex. The degree of complexity that we recognize is, to a considerable degree, a function of the speciation model that we are operating under. Although it is not possible to ex post facto determine with certainty the methods of speciation that occurred within Curimata, we can draw on available information on the ecology and behavior of the genus as a whole to evaluate the likelihood of alternative models. Curimata species undertake long distance spawning and feeding migrations, typically in larger schools, often through rapids and across other barriers. Although the species of Curimata have modifications specialized for utilizing detritus, the members of the different genera do not apparently specialize within that general food type. Such highly mobile populations and the lack of any apparent species specific specialization in diet or modes of reproduction make it difficult to conceive of Curimata species as undergoing either stasipatric or sympatric speciation. The limitation of Curimata species, and indeed all curimatids, to relatively low altitude drainage systems do, however, make them prime candidates for allopatric speciation resulting from vicariance between hydrographic systems. These factors in combination lead me to cast the following discussion within an allopatric speciation framework.

Only two species of Curimata do not have distributions overlapping to some degree with that of at least one congeneric. These are C. macrops (D) of the rivers of northeastern Brazil and C. mivartii (G) of the Río Magdalena system; river systems with overall limited curimatid faunas. The more typical condition for species of Curimata and one that is obvious in Figure 42 is the notable, often complete, sympatry between sister clades. That pattern is presumably indicative of secondary dispersal following allopatric spaciation. Such secondary sympatry obscures the underlying vicariance events that lead to speciation in the genus. Several less inclusive subunits of the genus do, however, demonstrate allopatric distributions, which are congruent with a vicariance biogeography model. These allow insight into both area relationships within the South American ichthyofauna, and into the relative timing of some of the speciation events in the Curimatidae.

The most discrete of the allopatric sister pair distributions within Curimata involves the three species of node 8 (G = mivartii; H = aspera; I = cerasina). Curimata mivartii of the Río Magdalena system is the sister species to the clade consisting of C. aspera of the western Amazon and C. cerasina of the Orinoco basin. This pattern of relationships is congruent with a hypothesis of a vicariance event between the Río Magdalena following temporally by a vicariance event between the Amazon and Orinoco basins. As noted by Vari (1988) this pattern of relationships and distributions is also interesting in indicating that the majority of the more encompassing speciation events of the genus (Figure 42, nodes 1–4) occurred prior to the Andean vicariance event, a pattern evidently general for the family. Additional allopatric distributions between sister species involve C. knerii (F) of the middle and upper Amazon and C. cyprinoides (G) of the Guianas and lower Amazon, and C. roseni (K) and C. inornata (L) of different subsections of the Amazon (see Figures 23 and 26) cannot, however, be similarly tied to specific geologic events at this time.

Returning to the intrageneric sympatry noted above, we find that there are repeated patterns of large scale sympatry between sequential sister groups (nodes 1–5) until we reach the terminal dichotomies within the genus (nodes 6–9). If the subdivision of the continent by the uplift of the Andes during the Miocene is correlated with the speciation event of node 8, then the speciation at nodes 1, 2, 3, and 4, which also demonstrate the largest degree of sympatry, would have predated the event. It is quite reasonable to assume that the sympatry between sister clades at those phyletically higher level nodes is correlated with the dramatic rearrangement of the hydrographic patterns in South America associated with the uplift of the Andes. The reorganization of the drainage patterns of the continent would have brought previously separated species into the same basins. The species could then have dispersed through those now conjoined systems. The lower level of sympatry or lack of distributional overlaps at nodes 6, 7, 8, and 9, in turn, could reflect the more stable hydrographic patterns that have existed following the final uplift of the Andes. Such hydrographic stability would have reduced the possibilities for secondary dispersal between drainage basins following speciation resulting from geological vicariance events.

Resumo

O gênero Curimata Bosc (1817) é definido como uma subunidade monofilética da família Curìmatìdae com base em várìas sinapomorfias nos arcos branquiais, complexo buco-faringeano e arco palatino. Modificações de natureza derivada nos arcos branquiais, aparelho hióide, aparelho opercular, osso maxilar, hiomandibular, arco palatino, supraneurais, primeiro pterigióforo proximal da nadadeira dorsal, pigmentaçaõ, forma do corpo, tamanho e forma das escamas, grau de desenvolvimento dos raios das nadadeiras dorsal e caudal, número de vértebras e morfometria reunem grupos de espécies dentro do gênero.

Curimata é redefinido e doze espécies são reconhecidas no gênero. Estas são: Curimata cyprinoides Linnaeus (1766), distribuída no baixo rio Orinoco, drenagens atlânticas das Guianas, baixo rio Amazonas e rio Tocantins; C. vittata Kner (1859), do sistema do rio Amazonas; C. aspera Günther (1868a), do alto rio Amazonas; C. knerii Steindachner (1877), do médio e alto rio Amazonas; C. mivartii Steindachner (1878), endêmica do rio Magdalena e rios associados; C. ocellata Eigenmann & Eigeinmann (1889), distribuída na bacia do rio Amazonas; C. macrops Eigenmann & Eigenmann (1889), conhecida somente da bacia do rio Parnaíba no nordeste do Brasil; C. cisandina (Allen, em Eigenmann & Allen, 1942), encontrada no sistema do rio Amazonas; C. cerasina Vari (1984b), endêmica do rio Orinoco; C. incompta Vari (1984b), da bacia do rio Orinoco; C. inornata, uma nova espécie do sistema do rio Amazonas; e C. roseni, uma nova espécie da bacia do rio Negro.

Verificou-se que a espécie Curimatus semitaeniatus de Steindachner (1917) foi baseada em um exemplar juvenil de Curimata ocellata Eigenmann & Eigenmann (1889), sendo por esse motivo colocada como um sinônimo da mesma. Curimata murieli, descrita por Allen (em Eigenmann & Allen, 1942), tem como holótipo um exemplar juvenil de Curimata vitata Kner (1859), sendo a primeira espécie considerada um sinônimo júnior da última. Lambepiedra alleni, de Fernández-Yépez (1948), é colocada como um sinônimo de Curimata cisandina Allen (em Eigenmann & Allen, 1942). Curimatus simulans, proposto por Eigenmann & Eigenmann (1889), é considerado como um sinônimo de Curimata aspera Günther (1868a). Salmo edentulus Bloch (1794), Charax planirostris Gray (1854), Curimatus schomburgkii Günther (1864) e Curimata copei Fowler (1906) são colocadas como sinônimos de Curimata cyprinoides Linnaeus (1766). Salmo immaculatus Linnaeus (1758), o nome mais antigo para a espécie aqui denominada Curimata cyprinoides, não é disponível (Comisso Internacional de Nomenclatura Zoológica, 1966).

Contrariamente a muitos estudos anteriores, Semitapicis de Eigenmann & Eigenmann (1889) é colocado como um sinônimo de Curimata Bosc (1817), em lugar de ser mantido como um gênero distinto. Os gêneros Acuticurimata Fowler (1941), Allenina Fernández-Yépez (1948), Bitricarinata Fernádez-Yépez (1948), Bondichthys Whitley (1953), Lambepiedra Fernández-Yépez (1948), Peltapleura Fowler (1906) e Stupens Whitley (1954) são também considerados como sinônimos de Curimata.

A zoogeografia histórica das espécies de Curimata é discutida, apresentando-se a hipótese de que a maior parte da divergência evolutiva a nível de espécie no gênero se deu antes do levantamento fīnal dos Andes. O grau de simpatria entre linhagens-irmãs dentro de Curimata indica que ocorreu uma significativa quantidade de dispersão pós-vicariância no gênero.