Florida Integrated Science Center
|
TAXONOMY AND SYNONYMY According to Nelson (1994), the group of teleostean fishes known as snakeheads is classified as follows:
Two genera are currently recognized as comprising the family Channidae. They are Channa (Scopoli, 1777; snakeheads of Asia, Malaysia, and Indonesia) and Parachanna (Teugels and Daget, 1984; African snakeheads). Generic synonyms of Channa include Channa Gronow, 1763, a nomen nudum; Ophicephalus Bloch, 1793, and its misspelled version Ophiocephalus; Bostrychoides Lacepede, 1801; and Philypnoides Bleeker, 1849. Synonyms of Parachanna are Ophiocephalus Gunther, 1861; Parophiocephalus Senna, 1924 (originally proposed as a subgenus, but preoccupied in the fish family Anabantidae by Parophiocephalus Popta, 1905); and Channa Scopoli, 1777. Synonyms of the 29 species of snakeheads described herein are included in the individual species accounts contained in the section “Species Accounts.” Myers and Shapovalov (1932) reviewed the status of the genera Ophicephalus and Channa, and they concluded that the practice of separating the two based on presence (Ophicephalus) or absence (Channa) of pelvic fins was invalid, based on specimens of C. gachua from India and one from Taiwan (introduced) that lacked pelvic fins. They placed Ophicephalus as a junior synonym of Channa. Five species of Channa lack pelvic fins. Vierke (1991b), Musikasinthorn (2000), Musikasinthorn and Taki (2001), and Zhang and others (2002) consider 29 species of this family as valid (table 1). Nevertheless, 87 species and 4 subspecies have been described (Eschmeyer, 1998, in part) and current taxonomy is in flux. Although many described species are now considered synonyms of recognized species, there are about 20 names that cannot be associated with valid taxa. The plethora of scientific names for snakeheads is in part due to the sometimes dramatic color changes that occur between early and late juvenile stages, and adult patterns, a factor then unknown and hence unrecognized by early taxonomists using color as one of the distinguishing characteristics. Moreover, some descriptions lack detail, illustrations, or type specimens that could assist in solving these taxonomic mysteries. Four new species have been described since 1990, another put into synonymy, and two removed from synonymy and recognized as valid during that same time period. A taxonomic revision of the family is being prepared (Prachya Musikasinthorn, personal commun., 2002) and will likely result in more species being recognized as valid, and new species will perhaps be described.
Table 1-Currently recognized species of the family Channidae - click to enlarge As is typical of fishes of foreign origin, there has been a history of different English common names utilized for snakeheads. It is not unusual to find dissimilar names used for juveniles and adults of the same species, particularly in the aquarium fish trade. Moreover, one can find several English common names in the scientific literature for the same species in different parts of its native range. This also is true for common names used by indigenous people for the same species. In India, for example, various common names for a single species are often used by people from diverse regions, states, or cultures. For purposes of this report, we have followed Robins and others (1991) in using common names for the two snakeheads they treated, selected names we felt appropriate primarily from those used in the aquarium fish trade, and have added proposed names for some species that lacked English common names (table 1). The common names are identified in table 1 and appear in bold type in the section “Species Accounts.” Accounts for each of these species are detailed in the section “Species Accounts,” which includes illustrations or photographs, source of original description, type specimens, synonyms, common name(s), native range, introduced range, size, habitat preference, temperature range, reproductive habits, feeding habits, characters, commercial importance in the United States, commercial importance in native range, and environmental concerns. Where known, the diploid chromosome number is included. Each account also contains a map showing native range and, where known, location or range of introductions. Literature citations for some synonyms are not included in the “References” section but can be found in Eschmeyer (1998) or on the Internet at http://www.calacademy.org/research/ichthyology/catalog/fishcatsearch.html. DESCRIPTION AND DISTINGUISHING CHARACTERISTICS Fishes of the family Channidae are commonly referred to as snakeheads (sometimes serpent-heads), primarily because of their somewhat elongated and cylindrical bodies, but particularly due to the presence of large scales on the head of most species, reminiscent of the large epidermal scales (cephalic plates) on the heads of snakes. Another snake-like feature is the somewhat flattened head with eyes located in a dorsolateral position on the anterior part of the head. Anterior nostrils are present and tubular. Dorsal and anal fins are elongated, and all fins are supported only by rays. A few species lack pelvic fins (Nelson, 1994; Berra, 2001). The caudal fin is rounded. The mouth is terminal and large with a protruding lower jaw, which is toothed, often containing canine-like teeth. The prevomer and palatines may or may not be toothed, depending on species. Scales are cycloid or ctenoid. All species possess paired suprabranchial chambers located behind and above the gills. The chambers in Channa are bordered by two plates, one from the epibranchial of the first gill arch and the other as an expansion of the hyomandibular. Those in Parachanna have a simple cavity not involving processes from the first epibranchial or hyomandibular. These chambers are not labyrinthic (Berg, 1947), but are lined with respiratory epithelium. All species occupy freshwater and a few can tolerate extremely low salinities. Illustrations or photographs of certain species of Channa appear in Nichols (1943), Munro (1955), Nakamura (1963), Mohsin and Ambak (1983), Masuda and others (1984), Lim and Ng (1990), Ng and Lim (1990), Lee and Ng (1991, 1994), Pethiyagoda (1991), Talwar and Jhingran (1992), Kottelat and others (1993), Jayaram (1999), and Kottelat (1998, 2001a). The three species of Parachanna were illustrated by Bonou and Teugels (1985), who provided a key for identification of Parachanna. But, there is no single key to identify all species of Channa, at least five of which appear to be species complexes rather than single, distinct species. Two larger snakehead species, the bullseye snakehead (Channa marulius) and emperor snakehead (C. marulioides), superficially resemble the native bowfin (Amia calva) in that all three are elongated fishes, have long dorsal fins, tubular nostrils, and an ocellus near the base of the upper part of the caudal fin. The bowfin, however, has its pelvic fins in an abdominal rather than thoracic or anterior-abdominal position, and the anal fin is not elongated. Moreover, the bowfin lacks a rosette of enlarged scales on top of its head. Other than this example, there are no native fishes in North America or within their native ranges with which snakeheads could be confused. Species and species complexes of the genus Channa are native from southeastern Iran and eastern Afghanistan eastward through Pakistan, India, southern Nepal, Bangladesh, Myanmar, Thailand, Laos, Malaysia, Sumatra, Indonesia, Vietnam, Korea, and China northward into Siberia (fig. 2). Of the currently recognized 26 species of Channa, 8 species and representatives of 4 species complexes occur in peninsular Malaysia, Sumatra, and/or Indonesia. Of the same 26 species, 13 species and 1 species complex are tropical to subtropical; members of 6 species and 2 species complexes are warm temperate to subtropical/tropical, 2 species complexes are cold temperate to subtropical/tropical, and 1 species is warm temperate to boreal and can live beneath ice in the northern part of its range. The three species of Parachanna are native to Africa and are tropical (fig. 2). Snakeheads are non-ostariophysan primary freshwater fishes (Mirza, 1975, 1995) and have little or no tolerance for seawater. Habitat preferences vary by species or species complex, with a majority occurring in streams and rivers. Others live in swamps, rice paddies, ponds, and ditches. All can tolerate hypoxic conditions because they are airbreathers from late juvenile stages. The pH range, where known, varies by species with one, the Bangka snakehead (Channa bankanensis), preferring highly acidic (pH 2.8-3.8) waters (Lee and Ng, 1991; Ng and Lim, 1991). At least three species are tolerant of a wide pH range: the dwarf snakehead (C. gachua), spotted snakehead (C. punctata), and chevron snakehead (C. striata) survived for 72 hours at pH levels ranging from 4.25 to 9.4 (Varma, 1979).
Figure 2-Native distribution of the family Channidae. click to enlarge Lydekker (1886) reported on fossilized skull bones of snakeheads from the Siwalik Hills, Himachal Pradesh, northern India. These fossils and additional material of Pontian age (early Pliocene) were described in more detail by Sahni and Khare (1977) who described Channa bhimachari, C. gregoryi, and C. romeri based on fossilized skull-bone material found in stream sediments. Comparing these skeletal elements to osteological preparations of recent snakeheads, they suggested that C. bhimachari was most closely related to C. striata and C. gregoryi to C. marulius. They were unable to determine a recent relative of C. romeri. Boeseman (1949) described skull-bone fossils of a snakehead from Pleistocene deposits from Trinil, central Java. Based on comparisons with osteological preparations of recent snakeheads, he concluded that these fossilized remains were of a species closely related to living Channa striata, and he named the species Ophicephalus palaeostriatus. Coupled with data presented by Lydekker (1886), Boeseman (1949), and Sahni and Khare (1977), there is convincing evidence that congeners of living snakeheads were present in Asia by and probably well before the Pleistocene. A majority of living snakeheads are native to southeastern Asia, with most species found from Myanmar eastward and southeastward through Malaysia, Sumatra, Java, and Borneo (Kalimantan). Boeseman (1949) confirmed that at least one species of what is now included in the genus Channa was present in Java in the Pleistocene. Jocano (1975) noted that during the Pleistocene, the Malay Peninsula, Sumatra, Borneo, and the “Sundas” to Palawan were connected by what is termed the “Sunda Shelf.” He described conditions as “a vast dryland covering 1.8 million square kilometers,” containing a “great river,” adding that many of the present river systems of Kalimantan, Sumatra, and surrounding localities were tributaries of that river. He also added that this “may explain the striking similarities of fish faunas in Sumatra, Borneo, and the Philippines,” the latter restricted to the Palawan region of the Philippines. From an evolutionary standpoint and considering a Pliocene to Pleistocene presence of channid fishes in Asia, it is likely that the “great river” of the Sunda Shelf began to dry during the late Pleistocene and speciation of snakeheads in the region was in advanced stages. This probably occurred before parts of that shelf became isolated as peninsulas and islands, with species evolving in differing kinds of habitats separated from their ancestral origin(s) by geographic and biological factors. Such an evolutionary scenario explains why, for example, species known from Sumatra also occur in the southern part of the Malay Peninsula and Kalimantan, Bangka, Billiton, perhaps Bali, and nearby areas. Moreover, isolation of such species “populations” between continental and insular ranges over geologic time has doubtless led to genetic and phenotypic differences that complicate taxonomic interpretations. Similar linkages between what are now continental areas and nearby islands (for example, Vietnam and China with Hainan) are known from the Pleistocene (Sterling and others, 2003). India and Sri Lanka are reported to have been connected as recently as 8,000 years ago (www.tamilinfo.org). Fossils of snakeheads have also been identified from post-Pleistocene deposits in the Sahara Desert (Van Neer, 1989). Banerjee and others (1988) suggested an origin of the family from the area of Yunnan Province, southern China, dating back to the Pliocene or earlier, but based their suggestion on ecological habitats currently occupied by snakeheads rather than from fossil evidence. A more accurate picture of where this family evolved and its ancestor(s) is yet to be determined. Most snakeheads build nests by clearing a generally circular area in aquatic vegetation, often weaving the removed vegetation around the centrally cleared area. This results in a vertical column of water surrounded by vegetation. Sometimes the surface of this column contains pieces of removed vegetation. One species complex (Channa punctata) prepares elaborate tunnels through vegetation leading into the nest column. In general, the male entwines his body around that of the female, with some species appearing to “dance” in the water column as eggs are released and fertilized (Breder and Rosen, 1966; Ng and Lim, 1990). Eggs are buoyant, due to a large oil droplet in the yolk mass, and rise to the surface where they are vigorously guarded by one or both parents. Some snakeheads in one species complex (C. gachua) and C. orientalis are reported to be mouthbrooders, with the male being the mouthbrooder of fertilized eggs and, later, fry in C. orientalis. Peter Ng (personal commun., 2002) suggested that C. asiatica may also be a mouthbrooder. Most snakeheads, however, are not mouthbrooders and one or both parents vigorously guard their young. One species (C. micropeltes) is reported to have attacked and, in some instances killed, humans who approached the mass of young (Kottelat and others, 1993). Thus, parental care, whether by guarding or mouthbrooding, is a behavioral characteristic of snakeheads. One might assume, based on reported spawning habits, that presence of vegetation is mandatory for successful spawning, but this is not the case. Wee (1982) cited Parameswaran and Murugesan (1976b) as having documented Channa gachua, C. marulius, and C. punctata spawning in ponds lacking vegetation. Alikunhi (1953) noted that C. striata is also known to spawn in the absence of vegetation. These observations, however, imply that other snakeheads are also capable of reproducing in waters lacking vegetation. Oocytes, when released from the female parent, are small, ranging from about 1 mm to slightly over 2 mm in diameter, depending on species. Fertilization takes place by the male releasing milt (sperm) on the oocytes as they emerge from the female. Development time to hatching varies with water temperature and, to a lesser extent, with the species involved. For example, hatching occurred in 54 hours at 16-26 oC and 30 hours at 28-33 oC in Channa punctata (Kahn, 1924). In the northern snakehead, C. argus, eggs hatch in 28 hours at 31 oC, 45 hours at 25 oC, and 120 hours at 18 oC. In general, newly hatched fry, depending on species, are about 3.0-3.5 mm in length. These suprabranchial chambers lie above the pharynx and gill arches, lateral to the otic chambers of the skull. In Channa, the chambers open into the pharynx through inhalant apertures. The chamber lining contains respiratory “islets” with vascular papillae. The chambers can be filled with air or water. In addition, in C. striata, there are also vascular papillae in the epithelium of the mouth and pharynx that can be utilized for respiration; these, however, can be retracted into depressions in the epithelium to prevent damage when feeding (Munshi and Hughes, 1992). Some channids, perhaps all, have a circadian rhythm in frequency of oxygen uptake. Channa marulius, for example, showed a peak in oxygen uptake at night. Channa striata and C. gachua peaked in early night hours, and C. punctata at dusk. These rhythms are attributed to evolution in swamp ecosystems (that is, the rhythm is a property of the ecosystem) (Munshi and Hughes, 1992). The number of species of snakeheads capable of overland migrations is unknown, but several display such behavior (Khin, 1948). These migrations often are assumed to be the result of fish relocating from drying habitats in search of those with water, perhaps driven by instinctive behavior for better feeding conditions, or both. Overland migrations likely apply to those species whose native range is subject to seasonal dry/wet (or monsoonal) conditions, which encompass much of western to southeastern Asia where the majority of snakehead species exist. The species of Channa most capable of overland migrations are those that are somewhat flattened ventrally (Peter Ng, personal commun., 2002). These include C. asiatica, C. gachua, C. micropeltes, C. melasoma, C. nox, C. orientalis, and C. striata. Even large C. micropeltes are capable of “crawling” in a sinuous motion on dry or wet land, although movement is slow (Peter Ng., personal commun., 2002). Those snakeheads with more rounded bodies (for example, C. argus, C. lucius, and C. maculata) have very limited ability to move on land except as young, and only when some water is present, as under mild flooded conditions. Liem (1987) noted that Channa, like airbreathing catfishes (Clarias and Heteropneustes), do not migrate on land to escape drying habitats, but burrow into mud to survive droughts. These species only migrate during or soon after heavy rains, allowing these fishes to invade new habitats, which permits a wider dispersal from more crowded environments (Liem, 1987).
Click Here to go to Associated Diseases and Parasites
|