ISSN 0936-9902 - Printed in Germany
Verlag Dr. Friedrich Pfeil, Wolfratshauser Str. 27, D-81379 München, Germany
Tel. + 49 - (0)89 - 74 28 27 0 – Fax + 49 - (0)89 - 72 42 77 2 – E-mail: info@pfeil-verlag.de – www.pfeil-verlag.de

A review of the South American cichlid genus Cichla,
with descriptions of nine new species
(Teleostei: Cichlidae).

Sven O. Kullander - Department of Vertebrate Zoology, Swedish Museum of Natural History, POB 50007, SE-104 05 Stockholm, Sweden.
E-mail: sven.kullander@nrm.se

Efrem J. G. Ferreira - Centro de Pesquisas de Biologia Acuatica, INPA, CP 478, 69083-000 Manaus, AM, Brazil.- E-mail: efrem@inpa.gov.br

Cichla mirianae

Geographical distribution:
Known from the upper Rio Tapajós drainage, in the Juruena and Teles Pires rivers, and from the middle and upper Rio Xingu drainage, in the Fresco, Batovi, Culuene
and Suiá-Missu rivers..

Etymology:
Named for Mirian Leal-Carvalho, who participated in the collection of part of the type series.

Notes:
Small specimens of C. mirianae display an indistinct continuous lateral band from snout to caudal peduncle (Fig. 37), and retain this band as adults displaying as irregular blackish blotches arranged in a narrow band connecting the three lateral ocellar blotches and continuing onto the caudal peduncle. This development of the colour pattern is similar to that of C. intermedia, but in
that species, the juvenile lateral band is wider and more prominent, and whitish spots covering body and head of C. mirianae are absent. Cichla intermedia lacks the prominent lateral ocelli of C. mirianae and has about 6 narrow vertical bars across the side. Adult C. mirianae are similar to C. orinocensis in possession of three prominent ocellated blotches along the side, but juvenile C. orinocensis do
not possess a continuous horizontal band along the side.

Local names: Tucunaré.

Geographical distribution:
Only three localities in the lower Rio Xingu drainage are verified by preserved material. The Balneário do Pedral is located on the Rio Xingu about 8 km from Altamira. The type locality, Cachoeira do Espelho is located 13 km south of Altamira. The third locality, Gorotire, is located on the lower course of the Rio Fresco.



Local names: Tucunaré.

Etymology:
The species epithet piquiti is a Tupi-Guarani word meaning striped, and is here used as a noun in apposition.

Local names: Recorded as tucunaré by Costi et al. (1977) and as tucunaré azul by Kelber (1999).

Notes:
Cichla piquiti is known as tucunaré azul among Brazilian recreational fishermen (Kelber, 1999), with reference to an overall light bluish impression of living specimens.

Cichla piquiti bears some resemblance to C. ocellaris and C. nigromaculata in the presence of five wide vertical bars on the side, but ocellar markings are absent from the side, and the vertical bars reach more ventrally on the side. Cichla piquiti can be distinguished from the sympatric C. kelberi by the absence of light spots on the anal fin, presence of bars 1a and 2a, absence of abdominal blotches, and absence of occipital bar. It may be confused at small sizes with C. pinima, which also occurs in the lower Rio Tocantins, but the two species appear to be only narrowly parapatric.

.

Geographical distribution:
Natural occurrences are recorded from the lower (Rio Itacaiunas, Marabá, Itupiranga), and upper Rio Tocantins (Rio Paranã), as well as in the Araguaia drainage (Cocalinho and Rio das Mortes).
Cichla piquiti is extensively stocked in reservoirs in the Paraná river drainage, and the States of Minas Gerais and São Paulo. Kelber (1999) lists tucunaré azul from the Itaipu, Porto Primavera, Jupiá, Três Irmãos, Ilha Solteria, São Simão, Jaguara, Estreito, Peixoto, Furnas, Avanhandava, and Promissão dams in the Rio Paraná drainage, the Tres Marias dam in the Rio São Francisco drainage, and the Serra da Mesa dam in the Rio Tocantins drainage, as well as from the Rio Piquiri in the upper Rio Paraguay drainage. The tucunaré azul corresponds to C. piquiti, according to photographs provided by D. Kelber. Photos of large adults were made available by D. Kelber from Tres Irmãos (Rio Tietê), collected in 1995, and Itumbiara (Rio Paranaíba). Preserved material is available from Embarcação and Bocaina, also in the Rio Paranaíba drainage.

Geographical distribution:
Rio Trombetas near Cachoeira Porteira, only upstream of the falls.

Etymology:
The species epithet thyrorus is from the Greek noun thyroros, meaning doorkeeper or porter, in reference to the geographical distribution in the Cachoeira Porteira, the Portuguese word porteira meaning gate or entrance.

Notes:
The Rio Trombetas apparently has a distinctive endemic fauna from the Cachoeira Porteira upstream. Kullander & Ferreira (2005) listed seven cichlid species that are known only from that region, and examples from other fish families.

They concluded that the lower rapids of the Trombetas may constitute a distribution barrier to some species, but at least three species are rheophilic. The lower rapids certainly are a barrier to upstream dispersal, as lowland Amazonian cichlid species are absent from the Rio Trombetas upstream from the Cachoeira Porteira, but present immediately below, e.g., Apistogramma pertensis listed by Kullander & Ferreira (2005). Among species of Cichla, this geographical separation is evident in C. thyrorus being present upstream and C. vazzoleri only in and downstream of the Cachoeira Porteira.

Local names: Tucunaré.

Geographical distribution:
Known only from the Rio Jari, in the region of Cachoeira Itacari and Cachoeira Santo Antonio.

Etymology:
Named with reference to the type locality, the Rio Jari; an adjective with alternative endings -inus and -inum.

Local names: Tucunaré.

Geographical distribution:
Lower Rio Tapajós, Rio Curuá-Una, lower Rio Xingu, lower Rio Tocantins, and Rio Capim; uncertain localities in southern Amapá and Rio Arapiuns, introduced in Rio Paraguaçu in eastern Brazil, and probably elsewhere in northeastern Brazil.

Etymology:
The species epithet pinima is a Tupi-Guarani adjective meaning spotted with white, and is here used as a noun in apposition.

Local names: We propose tucunaré pinima as Brazilian name for this species, selecting a name already in use.

Geographical distribution:
Cichla vazzoleri is recorded from the middle Rio Uatumã and
lower Rio Trombetas including but not above the Cachoeira Porteira, to the Oriximiná region at the mouth of the Rio Cuminá.

Etymology:
Named in honour of Gelso Vazzoler (1929-1987), collector of part of the type series, former head of the Departamento de Biologia Acuática, INPA, Manaus.

Notes:
Cichla vazzoleri is most similar to C. pinima. Refer to comparison under that species. Jégu et al. (1989) described Mylesinus paraschomburgkii, a rheophilic characid from the Rio Trombetas (Cachoeira Vira-Mundo and upstream) and Rio Uatumã (Cachoeira do Miriti and upstream).

This species appears to be restricted to rapids, and thus presents different dispersal options than the Cichla species. Although the distribution pattern of M. paraschomburgkii is similar to that of C. vazzoleri, which is present in both the
Uatumã and the Trombetas, C. vazzoleri, although collected from rapids in both rivers, is present in the lower Rio Trombetas chiefly in lowland conditions.

We are unaware of other fish species restricted to the Uatumã and lower Trombetas rivers, but we also cannot conclude that this is the total distribution of C. vazzoleri.

Geographical distribution:
Restricted to blackwater rivers and their tributaries, recorded from many localities along the Rio Negro in Brazil and Venezuela, many localities in the Rio Orinoco drainage, a few localities only in the Rio Branco drainage. Along the Rio Solimões-Amazonas recorded from Tefé, Rio Puraquequara, lower Rio Uatumã, Rio Preto da Eva, and Lago Saracá, all black-water habitats. Reported by Lowe-McConnell (1969) from the Rio Branco drainage in Guyana.
An adult specimen from the Rio Jamari (INPA 3493) and a small specimen from Lago Genipapo (INPA uncat.) on the Rio Aripuanã are questionably referred to C. temensis, but regrettably there is nearly no other Cichla material available from the Brazilian portion of the Madeira drainage to permit an understanding of the distribution of C. temensis in this region.

Local names:
Sarabiana (Natterer, in Heckel, 1840: 413, Rio Negro); tucunaré (Natterer in Heckel, 1840: 409, Boa Vista area); pintado (Schomburgk in Jardine, 1843:145, Rio Negro); Lucanari Grande (Schomburgk, in Jardine, 1843: 151, Rio
Negro); tucunaré sorubiana (Magalhães, 1931: 225, errata, Rio Negro); pavón pintado, pavón trucha, pavón venado (Román, 1981: 76, Venezuela); Pavón cinchado (large specimens; Román, 1981: 82).

Both the appelation pintado and sorubiana refer to the pimelodid catfish species Pseudoplatystoma fasciatum, known as surubi (Tupi-Guarani) or bagre pintado (Spanish).

Note:
Cichla temensis is the most elongate species in the genus, manifest above all at larger sizes, where it overlaps with C. vazzoleri but is more slender than C. pinima and C. thyrorus. The colour pattern and the ontogenetic development of principal dark markings, distinguish it from
all other species of Cichla.

Geographical distribution:
Specimens at hand are from the Rio Cinaruco in Apure state, and from the Rio Casiquiare (Rio Cunucuna, Rio Cataniapo), reflecting the fragmented range given by Machado-Allison (1971: fig. 13, also including the Rio Siapa drainage) (Fig. 9). Winemiller (2001: fig. 2) also included the lower Rio Caura in the distribution. There are no records from Colombia, although the species is probably
present also in that country.

Local names: Pavón real.

Discussion:
Species diversity. Despite being one of the most colourful, well-known, and most commercial of the genera of cichlids, of importance to artisanal as well as tourism based sport fishing, and among the largest fishes in the Neotropics, the taxonomic confusion in Cichla has remained considerable.
A major source of confusion has been the obvious acceptance in species of Cichla of much more than modest variability in colour pattern, both ontogenetic and between individuals, as well as a traditional recognition of just a few species. This has been possible because most authors have
only dealt with very few specimens or a restricted geographical region. In this paper, covering a wider geographic area, we have been able to organise phenotypic variation within the genus into spatially clustered units representing 15 distinct phylogenetic species.

The species diversity of Cichla is thus demonstrated to be much greater than indicated by current literature. Nevertheless, we believe the work on a revision of Cichla has just started. Our species accounts include references to samples that cannot be satisfactorily assigned, and some
species are still known only from a few specimens. Huge areas within the geographical range of the genus have not been sampled for fishes. We expect that there may likely exist between 20 and 30 species of Cichla, which require large series of specimens and more extensive sampling to be located and diagnosed. Hopefully, this revision can inspire more efforts in further revising the genus.