Wayne Leibel, 2003
Cichlid Room Companion

The Chaetobranchine Cichlids: Filter-Feeding Planktivores from South America

By , 2003. image

Classification: Species overview, South America.

(This article was originally published in Cichlid News, Oct-2001, pp. 6-15. It is here reproduced with the permission of author Wayne Leibel.
Chaetobranchus flavescens Chaetobranchus flavescens, the Combtail Basketmouth, sports a fringed caudal fin and filamentous extensions of the soft dorsal and anal fins. A young 6" male is shown here. Photo by Wayne Leibel.

South American cichlids of the genera Chaetobranchus and Chaetobranchopsis are "basket-mouths" of another flavor. Rather than being gape and-suck predators like the true basket-mouth (Acaronia) species, or even the three false basket-mouth species of the genus Caquetaia, these are gape-and-pump filter-feeding planktivores! Like the other basket-mouths, they all have protractile premaxillaries that allow them to 'gape' (extend out their mouths to create a vacuum tube; Pellegrin, 1904). But instead of gulping down fish and other large prey, the vacuum created is used to draw water through their mouths and over their gills (oral pumping). Plankton suspended in the water are strained and retained as food by filtering it through their very long and numerous, setiform (i.e., thin) gill rakers. This character is shared by species of both genera and is reflected in their respective generic names: chaet- is Greek for hair or bristle; branch- means gill (opsis- means 'looks like'). Unfortunately, this means that they are rather difficult aquarium subjects because of their specialized feeding requirements. But it also makes them interesting and challenging subjects for the advanced New World cichlidiot.

Nomenclature and Phylogeny of the Chaetobranchines

The genus Chaetobranchus was erected by Heckel in 1840 for his newly described species flavescens, which remains the type species of the genus. The name flavescens (meaning 'yellowish') refers to the bright yellow coloration festooning the opercular region of this fish. Somewhat later, Steindachner (1875) would describe a second species, called semifasciatus, which he also placed in Heckel's genus Chaetobranchus. Steindachner chose the species name, which means 'half-banded', in reference to the light vertical bars that extend from the dorsal fin downward about halfway to the belly. Several other colorational and morphological features easily distinguish the two species (see below). Both are characterized by very long, slender (setiform), numerous gill rakers (microbranchiospines), especially on the first gill arch, which is the basis for the genus name (Kullander and Nijssen, 1989). Kullander (1986; Kullander and Nijssen, 1989) hypothesized that this anatomical adaptation allowed these fishes to efficiently strain free-floating plankton from the water.

Chaetobranchus flavescens A wild (ca. 30 cm.) adult C. flavescens fresh from the Rio Negro (Brazil), showing the produced filamentous soft dorsal fin. The tail and anal fins have been "trimmed" by other fishes. Photo by Lee Finley.

These microbranchiospines have the same shape as those seen in the oscar, Astronotus; however, they are also found - uniquely among cichlids - on the lower pharyngeal tooth plate. Stiassny (1991) saw Astronotus and Chaetobranchus as near relatives (sister genera) and grouped them together as the chaetobranchines. However, despite the shared microbranchiospine shape, Regan (1906) believed Chaetobranchus to be very different from Astronotus on the strength of several other, unshared characters. Regan placed Chaetobranchus (and Chaetobranchopsis) together on their own basal branch after Cichla (the peacock 'basses') split from the ancestral Neotropical cichlid, but before Aequidens and Cichlasoma evolved from a more recent common ancestor. Regan (1906) believed Astronotus was descended from an Aequidens-type ancestor and was therefore not closely related to the chaetobranchine lineage. More recently, Kullander (1998) proposed the sub-family Astronotinae - made up of the tribes Astronotini and Chaetobranchini - as the basal (or most primitive, ancestral like) lineage of the Neotropical Cichlidae. In contrast, Farias and coauthors (2000), using molecular data (gene-sequencing) viewed the chaetobranchines as a sister-group to the geophagines (or eartheaters, comprising the genera Geophagus, Biotodoma, Mikrogeophagus, Teleocichla, Crenicichla, Apistogramma, Taeniacara, Guianacara, and Acarichthys), but are only distantly related to Astronotus, which together with Cichla, makes up the basal lineage of the family. As an aside, Regan (1906) saw the true basketmouths, Acaronia (then Acaropsis) species, as descended from an Aequidens ancestor along the lineage that gave rise to Nannacara, Geophagus, and Apistogramma. Thus Acaronia and the chaetobranchines are very distantly related at best. Likewise, Kullander (1998) proposed Acaronia as basal (near ancestral) to the subfamily Cichlasomatinae - the heroines (formerly included in the cichlasomines) and the cichlasomines (formerly acaras), and, again, relatively unrelated to the chaetobranchines. Finally, Farias et al. (2000) places Acaronia as the sister to the Laetacara-Bujurquina group of cichlasomines (acaras). It would seem that the 'basketmouth' character i.e., the protrusible mouth - is a good design shared by evolutionary convergence rather than common ancestry among several genera of South American cichlids, and that we as aquarists, by referring to them all as "basketmouths" confuse the real evolutionary story of their origins and (lack of) relatedness.

Chaetobranchus Species

In addition to C. flavescens, Heckel (1840), in the same monograph, described a second species, C. bruneus, which has since been synonymized with C. flavescens (Kullander, 1986). And apparently C. flavescens has been (re)described several times since (e.g., Centrarchus cyanopterus Jardine, 1843; Chromys ucayalensis Castelnau, 1855; Chaetobranchus robustus Günther, 1862; and Geophagus badiipinnis Cope, 1872; Kullander, 1986). One of the reasons for this confusing proliferation of names is the rather wide distribution enjoyed by C. flavescens; it is found throughout the Amazon basin (in Peru and Brazil), and occurs north to Guyana, Surinam and French Guiana, west to Venezuela and Colombia, and south to the Mato Grosso region of Brazil and Bolivia (Kullander, 1986).

The genus Chaetobranchus, as understood today (Kullander, 1986; Kullander and Nijssen, 1989), comprises two species, C. flavescens Heckel 1840 and C. semifasciatus Steindachner, 1875. The latter species has a more circumscribed distribution and is limited to the Brazilian Amazon; it has been collected near Manaus, Tefe, Obidos and Itacoatiara. As mentioned above, C. semifasciatus was named for its series of dark, vertical 'halfbands' that end midway down the flank (which C. favescens lacks); these are clearly drawn in the plate that illustrates Steindachner's description. By contrast, C. favescens has a prominent, dark blotch on the middle of the flank, which is lacking in C. semifasciatus. Unlike C. favescens, C. semifasciatus sports a conspicuous ocellus (black ringed in white) - much like that in Satanoperca daemon - on the caudal-fin base, which again is well-illustrated in the original description's accompanying plate. It also has 'scaly' vertical fins while C. flavescens has 'naked' fins (Kullander, 1986). Both species, however, have similar body shapes and a large, gaping, upturned, underslung mouth. And both are apparently food fishes throughout their respective distributions in Guyana and Brazil (Axelrod, 1993).

In short, there is no confusing these two fishes, though I have yet to see a live C. semifasciatus in the aquarium trade. Nevertheless, several photos of this species have appeared in the aquarium literature; for example, a color photo of several dozen dead adults in a marketplace in Manaus appears in both editions of Stawikowski and Werner (1988; 1998). There are also photos in Goldstein's (1973) Cichlids of the World and in Axelrod's (1993) Lexicon of Cichlids. In addition to the series of half-bands and the caudal ocellus, other photos and drawings (e.g., Steindachner, 1875; Fowler, 1954) show this species with a silvery base coloration and yellow throat (similar to but paler than in C. favescens), as well as caudal, dorsal, and anal fins with concentric black stripes on a reddish fin base, and just a hint of filaments in the dorsal fin. The best of these photos is the one already mentioned of dead fish with eyes glazed over piled up for sale in a market in Manaus. The only published "live" photo (in both Goldstein, 1973, and Axelrod, 1993) consists of one very stressed, nearly de-finned specimen, sporting just a hint of the diagnostic half-bars. Perhaps we'll get some live specimens for the aquarium soon. Steindachner (1875) gives a size range of his sample of C. semifasciatus of 16-29 cm (6.3-11.4 inches) in total length.

Chaetobranchus flavescens is a handsome, though understated, aquarium fish that has been imported irregularly. It is a vaguely iridescent (actually more metallic) silvery-gray fish punctuated by a blackish mid-lateral spot, the endpoint of a lighter and behaviorally-variable lateral stripe extending posteriorly from the gill cover. In addition, there is a series of very light vertical bars that extend from dorsum to ventrum and which, in certain behavioral states, darken on the ventral half of the fish, unlike in C. semifasciatus. The truly outstanding colorational feature of this fish is the bright yellow to orange throat and chest region (flavescens = yellow) that healthy adults exhibit. This is carried over to the cheeks and gill covers in the form of a gold iridescence and some iridescent blue, vermiform markings. The other dramatic feature of this fish is its fins; the tail fin develops long filamentous streamers (reminiscent of the combtail anabantid, Belontia signata) which is echoed by the multi-filamentous soft dorsal which extends back past the tail. These are rarely seen in newly imported fish, but develop with time and good care in the aquarium. Even Fowler (1954) figured C. flavescens with comblike extensions on the soft dorsal and anal (but not on the tail) though they are exaggerated. For this reason, I dubbed C. flavescens the "combtail basketmouth" (Leibel, 1993). All the unpaired fins are lightly spotted in reddish-brown, and the pelvics are edged in blue and red. Maximum reported adult size is 186.1 mm (7.3 inches) total length for Peruvia specimens (Kullander 1986) though I have heard of bigger examples (25 - 30 cm) taken in the Brazilian Rio Negro (Lee Finley, pers. comm.). Lowe-McConnell (1969) reports a maximum size of 280 mm (ca. 11 inches) standard length (i. e., minus the tail fin) for male specimens of Guyanese provenance. The largest female she recorded was 260 mm (10.2 inches) SL.

Chaetobranchus semifasciatus Chaetobranchus semitasciatus, drawing of the type specimen from Steindachner, 1875.

What little is known of the ecology and life history of Chaetobranchus species comes from fieldwork done by Rosemary Lowe-McConnell (1969) in Guyana. There, C. flavescens lives in savannah pools in the north. Based on their gill raker morphology and 'weak teeth', Lowe-McConnell (1969) deduced that C. flavescens is a plankton feeder (though this has never been confirmed by gut analysis; Knoppel 1970), and thus is restricted in its distribution to large bodies of waters where plankton are prevalent. Kullander (1986; Kullander and Nijssen, 1989) reports that C. flavescens is a denizen of turbid whitewater floodplain lakes in Peru and Surinam. The breeding season in Guyana occurred during the rainy season, and females with ripening gonads (typically smaller than males, maturing at 17 cm for females and 22 cm for males) could be seined from the pools in April, just prior to the rains. After the rains ended in September, juveniles 5-7.5 cm in length were caught. While Lowe-McConnell (1969) was able to report the breeding habits of several cichlids in her study, regrettably the turbid whitewater made observing C. flavescens impossible. She reports that nothing was learned about their breeding habits, but that "the lack of a caudal ocellus suggests it may be a mouthbrooder." Most probably they are not, but that awaits confirmation by captive spawning in clear water, as C. flavescens has yet to be spawned in the aquarium.

Chaetobranchopsis Species

The genus Chaetobranchopsis (meaning 'Chaetobranchus-like') was erected by Steindachner (1875) who originally coined Chaetobranchopsis as a name for a subgenus, but later elevated it to generic status. Therefore the name appears in the literature as Chaetobranchopsis (Steindachner 1875) with Steindachner's name in parentheses to indicate the change from the original designation (Goldstein, 1973). Chaetobranchopsis species share the setiform gill rakers and planktivorous habits of Chaetobranchus species, differing only in the higher number of hard anal fin rays (5-6 vs. 3), and the extensive scalation of their vertical fins (Kullander, 1986). Eigenmann and Bray (1894) in fact utilized the character "soft dorsal and anal naked vs. scaled" as the diagnostic separating the two genera. The genus was last revised by Regan (1906). Currently, there are two valid recognized species in the genus, C. australis and C. orbicularis, which are very similar in appearance. However, the former has five hard anal spines whereas the latter usually has six (versus three typical of Chaetobranchus species).

The type species of the genus is Chaetobranchopsis orbicularis, though first placed in Chaetobranchus by Steindachner in 1875. Chaetobranchus (Chaetobranchopsis) bitaeniatus Ahl, 1936 has proven to be a synonym of C. orbicularis though the name still commonly appears in aquarium literature (e.g., Sterba, 1962). The fish is easily recognized by the pair of dark, horizontal parallel bars on the flanks, one extending from the orbit to the top of the caudal peduncle and a second from the operculum at the level of the pectoral fin to the bottom of the caudal peduncle (hence the specific name bitaeniata or 'two lines'). Midway along the flank the upper stripe and, to a lesser degree, the lower stripe expand to form blotches.

Chaetobranchus orbicularis Chaetobranchopsis orbicularis in the aquarium of Lee Newman. Photo by Lee Newman.

However, these longitudinal stripes are behaviorally expressed, which may have been the reason for the independent descriptions of both C. orbicularis and C. bitaeniata; C. orbicularis is figured by Fowler (1954) as lacking the parallel stripes but with two spots on the flanks (orbicularis means 'a little circle or orbit'). In live fish the stripes fade leaving the two spots as the fish's behavior dictates. The body is otherwise metallic bronze or gold (with each scale with a glittering gold edge), and is somewhat squatter (deeper, strongly-compressed and higher-bodied) than the Chaetobranchus species. The vertical fins are yellowish with red bands and wavy lines, and the iris is bright red. There is no obvious sexual dimorphism. Chaetobranchopsis orbicularis hails from the whitewaters of central Amazonas and is rather widely distributed (Lowe-McConnell, 1991; Stawikowski and Werner 1998).

The second species, Chaetobranchopsis australis Eigenmann and Ward 1907 is found in subtropical western South America from Paraguay (Bahia Negra to Rio Paraguay) and Bolivia (Rio Guapore) eastward into Argentina with much the same distribution as Astronotus crassipinnis (the 'other' oscar). The specific name australis means 'southerly' referring to this distribution. This fish looks much like C. orbicularis although the lower longitudinal band is usually extended as black pigment over the ventrum, and larger adults (ca. 25 cm TL) grow grandly-produced dorsal and ventral fin filaments. Like C. orbicularis, C. australis has a metallic silvery to brassy sheen on its flanks, and the lower part of the head is yellowish to bright yellow. The iris is blood-red. There is a large black spot in the middle of the flank just below the upper lateral line. The upper angle of the gill cover has reddish to brownish spots. There is no apparent sexual dimorphism.

In The Aquarium

There is no question that these four chaetobranchine species are among the most obscure and rarely kept of South American cichlids. Although Axelrod (1993) writes that C. orbicularis was a popular aquarium fish during the 1960s, having been exported by Willi Schwartz from Manaus, where they are quite common along the Amazon's tributaries, an entry for them fails to appear in aquarium hobby literature of that era, including Exotic Tropical Fishes (Axelrod et al., 1962). But they are consistently in the German literature (e.g., Ladiges, 1934, 1954; Arnold and Ahl, 1936; Sterba, 1962). I personally don't remember seeing them except sporadically since the 1970s, and I would consider them uncommon in the hobby today, mostly because they are relatively plain, at least at small sizes. As such, they are of considerable interest to advanced cichlid aquarists looking for a challenge.

Of the four species, the combtail basketmouth (C. flavescens) is perhaps the species with the best chance of prospering and breeding in the aquarium. While they will vacuum filter-feed as their gill anatomy permits, particularly when young, they can be enticed to eat larger fare including freeze-dried krill and a variety of frozen foods. Some of the specimens of C. flavescens I've kept have even eaten pelleted foods. They are rather undemanding with regards to water chemistry, given their cosmopolitan distribution, which includes the white waters of Peru of average pH and hardness. I've gotten close to spawning these fish several times but have lost the adults before realizing this goal for several, rather unfortunate and lame reasons (e.g., heater stuck on and cooked them; overfed by tank-sitter during vacation; plastic tubing popped off valve and drained tank dry while I was at work; and so on). I guess I am not fated to spawn these fish! To the best of my knowledge they have not yet been spawned in captivity by anyone, including our German counterparts who have had them in their hobby since 1934; as such they represent a real and possible challenge. I have had courtship from young adults 15-17.5 cm in length, so there is probably no need to grow them out to full adult size which is 10-12 inches. Lowe-McConnell's (1969) observations in the wild in Guyana should provide useful information relevant to stimulating captive spawning. In addition, they should make great show fish as the combtail extensions of their caudal, anal, and dorsal fins grow continuously through maturity. As stated earlier, the other species in this genus, C. semifasciatus, has yet to be imported in any numbers; I have yet to see one alive in the aquarium trade.

The other genus, Chaetobranchopsis, has been represented in the hobby by the infrequently imported species C. australis and C. orbicularis, both of which I've had occasion to keep several times each-all with calamitous results. Chaetobranchopsis orbicularis was first imported into the hobby by Aquarium Hamburg (Germany) in 1934 (Sterba, 1962; Stawikowski and Werner, 1998), but never caught on in the American hobby. I've had C. australis of varying sizes and importations from Argentina over the past decade. These I have tried to feed newly-hatched brine shrimp and crushed (actually finely-milled) freeze-dried krill, which they dutifully and enthusiastically filtered from the water. They pause in mid-water and pump, taking repeated mouthfuls of water - by extending and retracting their basketmouths - over and through their gills, as gill covers open and close rhythmically and in reverse synchrony (i.e., mouth open with gill covers closed, then vice versa). Nevertheless, they all got thinner and thinner and eventually appeared to waste away terminally, the old "knife-edge" belly syndrome. They would mouth other foods - frozen, live or prepared - and the occasional individual would actually seem to be swallowing some, but they never lasted long. As my old friend John O'Malley and I were fond of saying of new imports (actually, somewhat grimly and never really fondly) "these fish are dead already and just don't know it yet."

My friend Lee Newman, curator of fishes at the Vancouver Aquarium, and a rather more careful aquarist then myself, was able to keep C. australis alive and healthy for about 6 months, but they took a lot of work. Although thin upon arrival at the wholesalers, they immediately fed greedily on live brine shrimp nauplii. They also ate frozen daphnia and 'zooplankton' (a wet food sold in jars in Canada). He had about a dozen (each ca. 7.5 cm TL) in a 33-gallon tank, and there was very little in the way of conspecific aggression. They were housed by themselves to avoid any feeding competition. Their abdomens did fill out during feeding, but that took a lot of nauplii. They were not really shy despite the relative sparseness of the tank furnishings and lack of dither fish. He kept them at 26-27°C, with a pH of 6.4-6.8. He had them for about 6 months (they did well the whole time) before deciding to 'donate' them to the Vancouver Aquarium due to tank space limitations. Unfortunately, they did not get the same care as he was able to provide at home and they slowly declined and died.

On the upside, the American aquarist and adventurer Dan Fromm, who brought back C. australis himself from Paraguay, reports that his captive group has been going strong now as a colony for several years in a standard 70-gallon tank with a submerged, water-logged tree stump covered with Java fern as a focal point-hiding place, and a breeding colony of splash tetras, Pyrrhulina australis, for company. He feeds them - and they eat -live adult brine shrimp and live Daphnia, which they pump-filter, and frozen bloodworms and flake food, which they simply swallow. He has had no breeding activity as of yet. In 1997 I was the happy recipient of a six-pack of 10 cm. C. orbicularis from Fred Kraus in Milwaukee (WI), and these fish ate pelleted foods (Tetrabits!) with gusto in Fred's tank. I kept them for two years but regrettably had no spawning activity. They continued to eat a diet of Tetrabits, frozen bloodworms and brine shrimp for me, but perhaps what was lacking was 'filterable' live foods like Daphnia or Artemia.


Hopefully one day someone will be able to get one of these species to spawn in the aquarium. Don Conkel turned his specimens of C. australis out into his Florida fishponds ten years ago and was able to get reproduction under these semi-natural conditions. In the aquarium is another matter, but we have some starting points courtesy of German collectors. Staeck and Linke (1995) found C. australis in the Rio Mamore drainage in Bolivia. The habitats where they were collected consisted of pond-like expansions of small streams and lagoons which were common on either side of the roads. All were typical (plankton-rich) whitewater biotopes with very turbid (muddy) but very soft, acidic water. They report a pH of 6.6, 1.5 degrees GH, 1° carbonate hardness, and a water temperature of 29°C. They were found living in a rich community with other cichlids including Cichlasoma boliviense, Crenicichla lepidota and Apistogramma staecki; characins including Pyrrhulina brevis, Charax gibbosus, Ctenobrycon spilurus and the predatory Hoplias malabaricus; knifefishes such as Gymnotus carapo and Eigenmannia virescens; and the catfishes Hoplosternum littorale and Parauchenipterus galeatus. They note that the collecting sites had huge numbers of small freshwater shrimp, and they suggest that these may play a key role in the diet of C. australis. Also, the water was covered with floating water hyacinth, Eichhornia crassipes, and the C. australis were collected just under the floating plants. In a slightly more comprehensive account of this species' distribution, Stawikowski and Werner (1998) cite a wider range of water parameters where this fish was taken in Bolivia: a pH range of 6-8; dGH range of 1-17; dKH range of 1-16, and water temperatures of 24-29°F, so the fish may be more forgiving than previously suspected. Nevertheless, these collecting data give us a starting point from which to create and then manipulate conditions necessary to stimulate spawning in the aquarium. While Stawikowski and Werner (1998) stress that C. australis is a filtering planktivore and should be offered Daphnia, Cyclops, Artemia, etc. in the aquarium, they also report observing these fish scraping algae off stones and wood, and filtering mulm.

I once said in print that the peacock basses, Cichla spp., were hopeless obligate piscivores that died in captivity unless fed live feeder fish and should be left in the wild (Leibel, 1993). Several aquarists (e.g., Milo Manden and Rick Perez of the greater Chicago area) have since proved me wrong by getting their specimens to eat krill and pelleted foods, and these have prospered and have even spawned successfully. So, I no longer a priori advise against keeping these cichlids, nor can I advise against attempting the chaetobranchines despite their rather specialized filter-feeding planktivorous diet in the wild. Some success in feeding both C. flavescens, and more impressively, C. australis and C. orbicularis has been achieved in the home aquarium suggesting that these are not necessarily obligate planktivores, and in the not too distant future I hope to be reading about the successful captive spawning of these little-known and under-appreciated South American filter-feeding cichlids. In fact, I still hope to do it myself.

Chaetobranchus orbicularis Chaetobranchopsis orbicularis "pump" filter-feeding with mouth extended. Photo by Lee Newman.

References Cited

  • Arnold, J.P., and E. Ahl; 1936; Fremdlandische Susswasserfische; Gustav Wenzel, Braunschweig, Germany. 592 pp.
  • Knoppel, H. A.; 1970; "Food of central Amazonian fishes"; Amazoniana, 2: 257-352.
  • Ladiges, W.; 1934; "Tropical Fishes for the Aquarium"; H. Koch, Hamburg, Germany. 236 pp.
  • Ladiges, W.; 1954. "Tropical Fishes/Tropische Fische (2nd ed)"; G. Wenzel, Braunschwig, Germany. 207 pp.
  • Leibel, W S.; 1993; "Fishkeeper's Guide to South American Cichlids"; Salamander/Tetra. 79 pp.
  • Lowe-McConnell, R. H.; 1969; "The cichlid fishes of Guyana, South America, with notes on their ecology and breeding behaviour"; Zool. J. Linn. Soc. 48:255-302.
  • Lowe-McConnell, R. H.; 1991; "Ecology of cichlids in South American and African waters, excluding the African Great Lakes"; In: M. H. A. Keenleyside, (ed.). Cichlid Fishes: Behaviour, Ecology and Evolution. Press, London. 378 pp.
  • Regan, C. T.; 1906; "A revision of the fishes of the South American Cichlid Genera Cichla, Chaetobranchus, and Chaetobranchopsis, with notes on the genera of American Cichlidae. Ann. Mag. Nat. Hist. (7) 17: 230-239.

References (17):


Leibel, Wayne. (Jul 06, 2003). "The Chaetobranchine Cichlids: Filter-Feeding Planktivores from South America". Cichlid Room Companion. Retrieved on Feb 04, 2023, from: