Cichlidophiles: Thinking Outside the Glass Box: Smitten by the Retroculus Bug

Author: Wayne S. Leibel

In last month’s “Cichlidophiles” (TFH August 2007), I highlighted the singular accomplishment of my good friend and successful cichlidophile Lee Newman of Vancouver Canada, who racked up the first recorded captive spawning of the chiselhead eartheater Satanoperca acuticeps back in 2001. This was no mean feat, since success had eluded aquarists on both sides of the Atlantic since the fish became available in the hobby at the turn of the Twentieth Century (it was described by Heckel in 1840). At least no mention of this accomplishment appeared in the modern eartheater bible by the European aquarist Weidner (2000, South American Eartheaters, Cichlid Press [English Translation]).

As I related in last month’s column, Lee’s success with these fish depended on a serendipitous observation of their “abnormal” feeding behavior relative to other Satanoperca species and his willingness to learn, experiment, and think outside of the glass box—strategies that have permitted the spawning of more than one “unspawnable” cichlid.

As aquarists desiring to spawn a particular fish, it is our duty to create the correct aquarium conditions tailored specifically to the target species in question. We must give them what they need. If we fail, they remain unspawnable. But if we have done our homework right, the unspawnable become another notch on our aquaristic belt. What follows is the story of how another unspawnable cichlid was successfully spawned for the first time in the world, just this year and again by Lee Newman—Retroculus xinguensis.

Retroculus, Ichthyologically Speaking

There are three species of Retroculus currently recognized by ichthyologists: R. lapidifer (Castelnau 1855), R. septentrionalis (Gosse 1971), and R. xinguensis (Gosse 1971). The first and oldest of these, R. lapidifer, was first described by F. de Castelnau in 1855 as Chromys lapidifera. (Indeed, de Castelnau described no fewer than 18 cichlids in the genus Chromys (1855, 1861), including several African tilapiines. The genus Retroculus was erected by Eigenmann and Bray in 1894 to contain their new species boulengeri, which they designated as the type for the genus (Eigenmann 1910) and which subsequently proved to be a synonym for C. lapidifera (Regan 1906).

Earlier, Steindachner (1875) had moved lapidifer to the genus Geophagus (Heckel 1840) due to the shared and diagnostic character of a lobed first gill arch having pronounced gill rakers. (The genus Chromys was abandoned for all of de Castelnau’s species due to a misspelling: it should have been spelled Chromis. Many cichlids were erroneously described in this genus of marine damselfish owing to their close systematic relationship and shared overall appearance and subsequently removed from it.)


The genus name Retroculus comes from the latin retro (backwards) on the back side and oculus (eye), and describes the relatively posterior position of the eye in the head as compared to other cichlids (May 2007, The species name lapidifer means “stone bearer,” which refers to an interesting behavior—and one of the keys, as it turns out, to their successful captive spawning—which was reported by de Castelnau in his original description. His type specimen, which he collected from the rapids of the Rio Araguaia in Brazil, was observed transporting small pebbles one by one in its mouth, which it used to construct a nest for its eggs (Gosse 1971).

Having collected a cichlid from the Oyapock River in French Guiana that resembled R. lapidifer, the Belgian ichthyologist J. P. Gosse (1971, Revision du Genre Retroculus [Castelnau, 1855], Bull. Inst. Royale Sci. Nat. Belg  47 (47): 1-14) set out to formally re-describe R. lapidifer (de Castelnau’s type specimen was apparently lost by the Museum of Paris) and to describe his new species, R. septentrionalis, and a third, R. xinguensis from the Rio Xingu. The apparent morphological similarities of Retroculus and eartheaters of the genus Geophagusprompted Steindachner (1875) to suggest that lapidifer’s closest relative was Geophagus (Satanopercaacuticeps. Indeed, in terms of overall appearance R. lapidifer does look like an eartheater, and it has been considered as such in the aquarium literature until only recently (I dubbed it the “goby eartheater” in 1991). Kullander (1998, “A Phylogeny and Classification of the South American Cichlidae,” in: Malabarba, et al. [eds], Phylogeny and Classification of Neotropical Fishes.

Edipucrs, Porto Alegre, Brazil) now places the Retroculus species group as a primitive, basal lineage to all other South American cichlids, and not at all closely related to the geophagines despite the convergent morphology.

R. lapidifer and R. xinguensis

Two of the three species, R. lapidifer and R. xinguensis, made their aquaristic debuts in the early 1990s when the Brazilian clearwater rivers Rio Tocantins and Rio Xingu yielded their treasure trove of wonderfully ornate (and incredibly expensive) loricariid catfish, like the zebra pleco Hypancistrus zebra. To make up the airfreight, cichlids from this region were also shipped out along with the plecos, introducing a number of exciting new species (new to both the hobby and to science) of eartheaters and pike cichlids. Some of these included eartheaters like the substrate-brooding “surinamensis”-type Geophagus argyrostictus and many beautiful geographic morphs of existing species (Geophagus altifrons) and other, as yet undescribed, species. And pike cichlids like Crenicichla compressiceps, C. cyclostoma, C. jegui, C. percna, C. phaiospilus, C. sp. Xingu I, II, III, etc., and ten or more species of goby “pikes” of the newly described genus Teleocichla Kullander 1988.

Arguably, the exponential upsurge in interest in pike cichlids over the past decade was driven primarily by these newly imported and very beautiful fish, some of which were described at the time of their importation. Others are still yet to be described. Along with these cichlids came R. lapidifer from the Rio Tocantins system (Rio das Mortes, Rio Araguaia) and R. xinguensis from the Rio Xingu system. Although differing on the basis of small changes in longitudinal line and caudal peduncle scale numbers, the main diagnostic difference between these two species is the presence (R. xinguensis) or absence (R. lapidifer) of a series of concentric blue iridescent stripes on the tailfin.

Retroculus in the Cichlid Hobby

As aquarium fish, both the two Retroculus species and the several Teleocichla species were underrepresented, primarily due to the difficulty in successful international shipping. Species of both genera are rheophilic and found almost exclusively in the rapids of the Xingu and Tocantins systems. As such, they require highly oxygenated and very clean water. Initial export at normal within-box cichlid densities led to the death of many, if not most. Successful shipment depended on lighter packing which resulted in fewer dead but higher prices: both sold at wholesale for prices between $25 and $50 apiece in 1990 U.S. dollars!

Retroculus Availability

Not surprisingly, mortality in the home aquarium was also typically the result of not providing highly oxygenated, low-nitrogen-waste environments. It quickly became clear that powerheads and/or heavy aeration and trickle filters were required to keep these fish happy and healthy. The same held true for many of the larger, more spectacular “pleco”-type catfish also taken from the rapids. I remember many of these catfish fetching upwards of $100 each wholesale for larger specimens in the early 1990s. (I even bought one large, outrageously red specimen myself for $125, but that was nothing!) These days, though still only sporadically available, the price of wild Retroculus has pretty much stayed the same, which means they cost relatively much less in 2007 U.S. dollars. Nevertheless, they are still uncommon if not rare in the American cichlid hobby.

My Early Trouble

I must admit to my own disappointments and failures with respect to these rheophilic cichlids (and that expensive pleco as well). My early disasters were a byproduct of these early importations: what fish survived shipping to actually be sold were already severely compromised. I vividly remember having finally “scored” some Teleocichla—three fish—all of which died in their individual bags on the two-hour trip home from the wholesaler (who graciously gave them to me, his only survivors of a boxful) to my home. I was devastated to say the least, having waited anxiously for months to get some. I have had Retroculus xinguensis twice, once in the early 1990s and again about eight years ago, both times with similar results—they lived for nearly a year and then bloated and died. This was despite understanding their requirement of moving, clean, highly oxygenated water, and my attempt—although clearly not sufficient—to provide these conditions. Or perhaps it was something else that was lacking, diet perhaps. I don’t know. I have not tried them again since, though they have been made available to me, on occasion. Like S. daemon (see my July 2007 column), these are “dead fish swimming” for me and my system—so what’s the point?

But any cichlidophile who has seen them, or even photos of them, is inexorably drawn to them, particularly if your “thing” is South American eartheaters. They are a “must have”! Thus was my friend and fellow cichlidophile Lee Newman, Curator of Tropical Waters at the Vancouver Aquarium, smitten with the Retroculus bug. It took him years to finally get some live specimens, but once he did you can bet the “unspawnable” label dissolved away under his magic touch, much as was the case for Satanoperca acuticeps. Actually, it wasn’t “magic” so much as it was Lee’s willingness to research, learn, and experiment—to think outside of the glass box. And the secret to his success with this fish will become apparent in next month’s column. Stayed tuned, cichlidophiles!

See the full article on TFH Digital