Marine coastal, brackish, and freshwater, Indo-West Pacific (Africa to Japan, Fiji, and Samoa).
Body oblong to oblong-ovate, somewhat compressed; opercle with two spines, lower spine longer; dorsal fin with notch, 11-14 spines and 8-14 soft rays, spinous portion depressible into a groove formed by a sheath of scales; anal fin with three spines and 7-12 soft rays; pelvic fins inserted distinctly behind base of pectoral fins, with one spine and five soft rays; caudal fin rounded, truncate, or emarginate; lateral line continuous and extending onto caudal fin; vomer and palatines of most species lacking teeth; six branchiostegal rays; paired extrinsic swim-bladder muscles arising from rear of skull or posttemporal and inserting on anterodorsal surface of the anterior chamber of the swim bladder (employed for sound production—a few other perciforms have sonic muscles,
but they differ in position); swim bladder transversely divided (Vari, 1978, gives details on these features of the swim bladder which are unique within the per-ciforms); 25-27 vertebrae. Maximum length about 80 cm.
Much work on this family has been done by R. P Vari. Nelson (1984, 1994) used the family name spelling Teraponidae; I now follow Eschmeyer (1998) in using spelling Terapontidae. Most of the freshwater species occur in Australia and in New Guinea.
Sixteen genera (e.g., Amniataba, Bidyanus, Hephaestus, Leiopotherapon, Mesopristes, Pelates (synonym Helotes), Pingalla, Scortum,, Syncomistes, Terapon, and Variichthys (a replacement name for Varia) with about 48 species (e.g., Yoshino et al., 2002).
Family KUHLIIDAE (400)—flagtails. Marine, brackish, and freshwater; Indo-Pacific.
Dorsal and anal fins each with a well-developed scaly sheath; dorsal fin deeply notched, with 10 spines and 9-13 soft rays; anal fin with three spines and 9-13 soft rays; no scaly pelvic axillary process; opercle with two spines; 25 vertebrae. Color generally silvery, often with dark markings on the caudal fin. Maximum length up to 50 cm. The three nannopercine genera formerly placed in this family are now placed in the Percichthyidae. The most wide-ranging species and the only one to occur in the Americas is the coastal Kuhlia mugil (synonym taeniura), which extends from Africa to the tropical eastern Pacific, sometimes occurring in freshwater. The common name for the family in Hawaii, aholeholes, is also used elsewhere.
One genus, Kuhlia, with about 10 species; most of the species are marine and brackish water but one, K. rupestris, occurs primarily in freshwater, and several others extend up rivers in continental areas and on oceanic islands (e.g., Tahiti) (e.g., Randall and Randall, 2001.
Family OPLEGNATHIDAE (401)—knifejaws. Marine; Indo-Pacific (South Africa, Japan, southern half of Australia including Tasmania, Hawaiian archipelago, Galapagos, and Peru).
Teeth in adult united to form a parrotlike beak (as in Scaridae, but with usual percoid type of pharyngeal dentition, capable of crushing barnacle shells and
sea urchins); spinous dorsal fin low in adults, basically as high as soft dorsal and continuous with it in juveniles; dorsal fin with 11 or 12 spines and 11-22 soft rays; anal fin with three spines and 11-16 soft rays; scales very small (unlike in scarids where they are large). Maximum length about 0.9 m.
One genus, Oplegnathus, with about seven species.
Superfamily Cirrhitoidea. Pelvics rather far behind pectorals; lower 5-8 rays of pectorals unbranched, usually thickened, and sometimes separate from one another; anal fin usually with three spines. In the provisional arrangement of cirrhitoid family relationships, Greenwood (1995) showed that the Cirrhitidae was most plesiomorphic, followed by the Chironemidae, with the Aplodactylidae, Cheilodactylidae, and Latridae forming an unresolved trichotomy. Changes in the family classification as shown here and the need for further study are also suggested in the study of Burridge and Smolenski (2004). Their molecular study suggested that the Latridae should be expanded to encompass some cheilodactylids, and further studies are desirable to understand the cladistic relationships of all genera in this superfamily.
Five families, 23 genera, and about 73 species.
Family CIRRHITIDAE (402)—hawkfishes. Marine; tropical western and eastern Atlantic, Indian, and Pacific (majority are Indo-Pacific).
Dorsal fin continuous with 10 spines and 11-17 soft rays; pectoral rays 14, the lower five to seven unbranched and membrane incised; cirri on interspinous membrane; anal fin soft rays 5-7; scales cycloid or ctenoid; swim bladder absent; vertebrae 26. Maximum length about 55 cm.
Hawkfishes are usually small and richly colored fishes that live in rocky and coral habitats. They have many features in common with the scorpaenids.
Twelve genera (e.g., Amblycirrhitus, Cirrhitichthys, Cirrhitops, Cirrhitus, Isocirrhitus, Notocirrhitus, Oxycirrhites, and Paracirrhites) with about 33 species (Randall, 2001a).
Family CHIRONEMIDAE (403)—kelpfishes. Marine; coastal Australia and New Zealand.
Dorsal fin with 14-16 spines and 15-21 soft rays; anal fin soft rays 6-8; vomer with teeth, palatines without teeth; jaw teeth conical or villiform. Maximum length about 40 cm.
Two genera, Chironemus (about 4) and Threpterius (1), with about five species.
Family APLODACTYLIDAE (Haplodactylidae) (404)—marblefishes. Coastal marine; southern Australia, New Zealand, Peru, and Chile.
Dorsal fin with 14-23 spines and 16-21 soft rays; anal fin soft rays 6-8; vomer with teeth; jaw teeth incisiform, lanceolate, or tricuspid.
One genus, Aplodactylus (synonyms Crinodus and Dactylosargus), and about five species (Russell, 2000; Burridge, 2000).
Family CHEILODACTYLIDAE (405)—morwongs. Marine; parts of the Southern Hemisphere in Atlantic, Indian, and Pacific, and in the Northern Hemisphere off coasts of China and Japan and the Hawaiian Islands.
Dorsal fin continuous, may be almost separated, with 14-22 spines and 19-39 soft rays; anal fin with three spines (third may be difficult to detect) and 7-19 soft rays; vomer and palatines toothless; lower four to seven pectoral rays in adults usually thickened, elongated, and detached (free of rest of fin); usually 24 vertebrae. Maximum length about 1.0 m. The only species in the Northern Hemisphere belongs to the subgenus Goniistius of the genus Cheilodactylus, a taxon with an antitropical distribution. The last major revision was by J. E. Randall in 1983.
Five genera, Acantholatris, Cheilodactylus (synonym Goniistius), Chirodactylus, Nemadactylus, and Sciaenoides, with about 22 species (e.g., Burridge, 2004).
Family LATRIDAE (406)—trumpeters. Marine; coastal southern Australia, New Zealand, Chile, and in the southern Atlantic.
Dorsal fin with 14-24 spines and 23-40 soft rays; anal fin with 18-35 soft rays; vomer with or without teeth. Trumpeters form an important sport fishery and are known for their fine taste.
Three genera, Latridopsis, Latris, and Mendosoma, with about eight species.
Superfamily Cepoloidea. It would be desirable to demonstrate in a phyloge-netic study whether or not the formerly recognized Owstoniidae and Cepolidae (e.g., Nelson, 1984) represent monophyletic lineages within the Cepoloidea and, if so, to evaluate whether or not they appear as distinct from one another as do other percoid families.
Family CEPOLIDAE (407)—bandfishes. Marine; eastern Atlantic (off Europe and in Mediterranean) and Indo-West Pacific (including New Zealand).
Dorsal fin continuous, with 0-4 spines (usually three); anal fin with 0-2 spines; vomer and palatine toothless; single postcleithrum; six branchiostegal rays; lateral line along base of dorsal fin; epipleural ribs in some trunk vertebrae fused proximally to corresponding pleural ribs; body color generally red or pink. Maximum length 70 cm, attained in Cepola rubescens of the northeastern Atlantic and Mediterranean.
Four genera with 19 species (e.g., Smith-Vaniz in Smith and Heemstra, 1986). Two subfamilies are provisionally recognized.
Subfamily Cepolinae. Highly compressed elongate body gradually tapering to the tail; dorsal and anal fins very long, each with over 65 rays, and connected with caudal fin; minute scales; about 65-80 vertebrae. Two genera, Acanthocepola (4) and Cepola (3), with about seven species. Shown in upper figure. Cepola haastii is the only cepolid from New Zealand.
Subfamily Owstoniinae. Elongate caudal fin; dorsal and anal fins each with fewer than 32 rays; 28-30 vertebrae. Deepwater. Two genera, Owstonia (11) and Pseudocepola (one), with 12 species. Shown in lower figure.
Family ELASSOMATIDAE (ELASSOMIDAE) (408)—pygmy sunfishes. Freshwater; eastern United States (including the Mississippi River Valley).
Infraorbitals (except lachrymal), basisphenoid, and endopterygoid absent; branchiostegal rays five; gill membranes broadly united across isthmus; den-tary and angular not penetrated by lateral line; no lateral line on body; caudal fin rounded; cycloid scales; dorsal fin with 2-5 spines and 8-13 soft rays; anal fin with three spines and 4-8 soft rays. Maximum length 4.5 cm.
Pygmy sunfishes were often included in the Centrarchidae in literature of the 1980's and earlier (e.g., Nelson, 1976, 1984) or placed in their own family but regarded as related to the Centrarchidae. Evidence published by G. D. Johnson in 1984 showed that they are probably not neotenic centrarchids, and that their affinities may even lie outside the Percoidei (supported in Johnson, 1993). Furthermore, Johnson and Patterson (1993) found that they share some derived features with the mugilomorphs, atherinomorphs, gas-terosteiforms, and synbranchiforms, and they included the elassomatids within a group comprising these taxa and termed by them the Smegmamorpha. G. D. Johnson and V. G. Springer in a 1997 conference talk presented evidence that they are closely related to the Gasterosteidae, but they later concluded that the matter remains unresolved (Spinger and Johnson, 2004:117) and placed them in their own order, Elassomatiformes, between Synbranchiformes and Mugilomorpha (Spinger and Johnson, 2004:124). The limited molecular evidence of Jones and Quattro (1999) is more consistent with a relationship between Elassoma and Centrarchidae plus Moronidae than between Elassoma and atherinomorphs or cichlids. Roe et al. (2002) did not find support for a relationship to either centrarchids or gasterosteids, but stated that a sister relationship with Centrarchidae could not be rejected. While I have removed pygmy sunfishes from the Percoidei here, I have retained them in the Perciformes, as in Nelson (1994), pending a clearer resolution of their relationships.
Gilbert (1998) gave a type catalogue of recent and fossil taxa.
One genus, Elassoma, with six species (e.g., Nelson et al., 2004; Gilbert, 2004).
Suborder Labroidei. The recognition of monophyly for this clade is based on characters in the pharyngeal region, particularly in the pharyngeal jaws, which are specialized for food processing. For example, i) the fifth cerato-branchials are united or fused to form a single lower pharyngeal jaw, suspended in a muscle sling (single or double), and ii) the upper pharyngeal jaw articulates by means of a diarthroses with the basicranium (see Stiassny and Jensen, 1987, for a detailed and critical discussion). However, as noted by Stiassny and Jensen (1987) and Johnson (1993), there is a high degree of homoplasy involved, and corroborative evidence from characters independent of the pharyngeal area is lacking. I echo the advice of Johnson (1993), considered especially relevant for this group given the nature of the synapo-morphies and the great interest by nonsystematists in the families, that workers seeking to understand the historical origins of various traits (e.g., behavioral) not regard the monophyly of this taxon, with cichlids being the postulated sister-group of the remaining families, as dogma.
The last three families labrids, odacids, and scarids form a monophyletic group and are combined into one family, the Labridae, by some past workers. However, while not necessarily arguing that each family might be mono-phyletic, I prefer, as in Nelson (1994) and most specialists, to recognize each until phylogenetic information is available to suggest otherwise (however, a cladistic analysis is needed of all genera to follow up on the statement in Bellwood, 1994:54, supporting the 1979 finding in the PhD dissertation of M. F. Gomon, that the Scaridae, the labrid Pseudodax, and the labrid hyp-sigenyine genera may form a monophyletic assemblage. Springer and Johnson (2004) recognize these three families in the superfamily Labroidea: while this is quite appropriate, I do not do so because of the uncertainty in how to rank the other three families (e.g., in three separate superfamilies).
Most species of labrids and scarids are protogynous, that is, they can change their sex from female to male; there is a wide diversity of color patterns associated with sex and size. Males may be primary (not capable of sex change) or secondary (resulting from a sex change of a female). Various populations may consist of only secondary males (monandry) or primary and secondary males (diandry). In coloration, individuals with a color pattern characteristic of small adults are said to be in the initial phase, while those having a color pattern characteristic of the largest males are in the terminal phase. Sexual dichromatism is common and refers to the situation where terminal-phase males have a different color pattern from females.
Six families with about 235 genera and roughly 2,274 species. About 1,330 species, all being cichlids, are confined to freshwater (an embiotocid is also virtually confined to freshwater). Fossils include the Eocene Tortonesia (Tortonesidae), thought to be related to the Pomacentridae, with the uppermost branched principal ray of the caudal fin greatly elongated and 25 or 26 vertebrae (described by L. Sorbini and coauthors in 1990), and the Eocene labrids Eocoris and Phyllopharyngodon. The African cichlid fossil record, known as early as the Eocene, was discussed in Murray (2000) and Stewart (2001).
Family CICHLIDAE (409)—cichlids. Freshwater and occasionally brackish water; Central and South America (one species extending north to Texas), West Indies, Africa, Madagascar, Israel, Syria, coastal India, and Sri Lanka.
Single nostril on each side; lateral line interrupted, generally 20-50 scales in lateral lines but number may exceed 100; generally 7-25 spines and 5-30 soft rays in dorsal fin and 3-15 spines (3 in the majority of species) and 4-15 soft rays in the anal fin. (Etroplus has about 12-15 anal spines, but most other species exceeding three spines have 4-9; a few cichlids may have more than 30 soft rays in the anal fin.) No subocular shelf. Maximum length about 80 cm, attained in Boulengerochromis microlepis of Lake Tanganyika.
As in many families, there is much variability in body shape between some species. Most cichlids have a moderately deep and compressed body similar to Cichlasoma shown in the figure. However, the body can be disc shaped and have extremely high, saillike fins, as in Pterophyllum (angelfishes), or low fins, as in Symphysodon (discus fishes); it can also be elongate, as in Crenicichla (pike cichlids).
Cichlids form an important group of relatively large and often colorful aquarium fishes. Many color patterns have been developed through selective breeding in some of the species for the aquarium trade. Keep in mind that the common names for the species of Pterophyllum, the angelfishes, can be confused with the pomacanthids, which are also known as angelfishes; the first group is freshwater and the latter is marine, and, where confusion could arise, refer to them as either the freshwater angelfishes or the marine angelfishes. There is an extensive aquaculture for several species of tilapia, especially for Oreochromis niloticus.
Species of the family have highly organized breeding activities (see Keenleyside, 1991). Two general forms of parental care may be recognized: i) mouthbrooders, which are usually polygamous and usually only the female carries the fertilized eggs and newly hatched eggs in the mouth; and ii) sub-stratebrooders, which are usually monogamous and both sexes may care for the eggs. A few species combine both methods, e.g., eggs are laid and cared for on the substrate, but the newly hatched young are carried in the parent's mouth. Mouthbrooding or oral incubation is common and appears to have arisen independently in several groups of African cichlids but is known from
only a relatively few species in Central and South America. Female discus fish secrete a whitish milklike substance from the skin to "nurse" their young.
Cichlids have attracted much attention in evolutionary biology because of the existence of species flocks in Africa. A wealth of information on the biology, adaptive radiation, and speciation of African cichlids exists. Endemic cich-lids make up most of the fish fauna in the three African lakes that contain the most species of fish of any lake in the world, namely, Lake Malawi, Lake Victoria, and Lake Tanganyika. These cichlids exhibit a vast diversity of feeding habits, including species specialized to eat the scales of other fishes. Berra (2001) summarized much of what is known on the above subjects (see also Keenleyside,1991). Much of the recent major contributions to the systematics and taxonomy of cichlids in Central and South America has been made by S. O. Kullander, and in Africa by M. L. J. Stiassny and by T. Takahashi. Their works should be consulted for details and only a few references are given here.
Cichlids are the most species-rich non-Ostariophysan family in freshwaters worldwide (Kullander, 2003). Several species of tilapias (species of Tilapia, Sarotherodon, and Oreochromis) are euryhaline and can disperse along some brackish coastlines between rivers. In Africa there are at least 900 species (with an estimated total of 1,300 when all are discovered and described) (see also Turner et al., 2001), four species in the Jordan Valley in the Middle East, one in Iran, three in India and Sri Lanka (also in brackish water), 17 on Madagascar (some also in brackish water), four in Cuba and Hispaniola (some in brackish water), 111 in North and Central America, and 291 in South America (Kullander, 2003, with figures updated from Kullander, 1998). There are thus at least an estimated 1,300 valid species of cichlids (a figure also given in Nelson, 1994), with several hundred more probably existing (a number of recent species described primarily in Africa gives an estimated 1,350). Several authors have discussed whether or not the cichlids in various continental areas form a monophyletic group and this will not be explored here (but for a discussion of cichlid biogeography, see Chakrabarty, 2004).
There is much active work on the taxonomy and systematics of cichlids (e.g., Kullander, 1998; Stauffer et al., 1997; Takahashi, 2002, 2003, 2004; Koblmuller et al., 2004). While progress is being made, there are many areas of disagreement and agreement, but we still lack a comprehensive system of assigning species to monophyletic genera. We are by no means at a point where a classification of the genera can be given despite attempts by Kullander (1998) and others. There is not complete agreement on the genera that should be recognized in this family. Because of this, in the list of common and scientific names of fishes from the United States, Canada, and Mexico, Nelson et al. (2004), pending a phylogenetic study of all species involved, recognized many species that probably belong in other genera, in a broad genus Cichlasoma. They also gave the genera into which they are placed in Kullander (2003).
Roughly about 112 genera: for example, the phylogenetically primitive Paratilapia and Ptychochromis of Madagascar; Acaronia, Aequidens, Amphilophus, Apistogramma, Astronotus, Cichla, Cichlasoma, Crenicichla, Geophagus, Gymnogeophagus, Herichthys, Heros, Parachromis, Pterophyllum, Retroculus,
Symphysodon, Theraps, Thorichthys, Uaru, and Vieja from Central and South America; Alcolapia, Chromidotilapia, Haplochromis, Hemichromis, Heterochromis, Julidochromis, Lamprologus, Oreochromis, Pelmatochromis, Pelvicachromis, Pseudocrenilabrus, Pseudotropheus, Sarotherodon (ranges to Syria), Teleogramma, Tilapia, Trematocara, and Tylochromis from Africa; Tristramella, endemic to Sea of Galilee, Israel; Iranocichla, endemic to southern Iran; and Etroplus, from India and Sri Lanka. As stated in Nelson (1994), it is almost meaningless to attempt to give a number of recognized valid species because estimates vary widely due to differing definitions of a species (with special problems in the species flocks of Africa, not to mention the possible loss of undescribed species from the introduction of a species of Lates in Lake Victoria). In addition, there undoubtedly are many species yet to be described and perhaps many yet to be placed in synonymy with others. However, because an objective of this book is to provide the number of species currently recognized as valid in each family, I give the figure of 1,350. Nelson (1994) and Kullander (2003) estimated at least 1,300; the slight increase estimated here is based largely on species described recently from Africa.
Was this article helpful?