Cichlidophiles: She's So Heavy

Author: Wayne S. Leibel PhD

By way of honoring the late Dr. George W. Barlow, whose death this past July surprised and saddened cichlidophiles the world over, I’d like to share the results of some research he started just a few years ago. Please see last month’s column (TFH November 2007) for a short biography and tribute to one of Cichlidom’s true greats.

George’s Last Research Paper

As professor emeritus, George retained office and lab space and continued to write and do research on cichlids, often with Berkeley undergraduates, until the time of his death. The study I wish to share with you was done during this time frame—actually in 2004—with the assistance of a then undergraduate, Jonathan Lee, whom he mentored. The work was published in October of 2005 (Barlow, G. W. and J. S. Lee, 2005. “Sex-Reversed Dominance and Aggression in the Cichlid Fish Julidochromis marlieri.” Ann. Zool. Fennici 42:477–483).

To the best of my knowledge this is the last formal research paper George published. George continued this work on Julidochromis with another Berkeley undergraduate, Ray Engezer. He describes this on his website, but a formal account of that further work remains unpublished.

Devils Red and Otherwise

Much of George’s earlier work was done on the role of aggression and dominance in the mate choice and breeding behavior of the Midas cichlid, Amphilophus (“Cichlasoma”) citrinellus (GÜnther 1864) from Lake Nicaragua (and more widely in Central America). Many hobbyists know this fish by its other hobby name, the red devil, which is also applied to a related fish from the same lake, Amphilophus (“Cichlasoma”) labiatus (GÜnther 1864), the big-lipped red devil. A third closely related cichlid, the arrow cichlid Amphilophus zaliosus, is also found living together with citrinellus and labiatus in Lake Nicaragua and its satellite lakes; it was formally described by Barlow in 1976. More recently, several additional distinct species from the Midas cichlid complex have been described—e.g., A. amarillo, sagittae, xiloaensis (Stauffer & McKaye 2002). Aquarists know them to be large, very aggressive (hence the name “devil”) show cichlids with variable coloration (gray-barred to golden to bright orange-red, or even piebald), which has made them quite popular in the cichlid hobby for many years.

George’s work, which details the complex social life and breeding behavior of these remarkable cichlids in nature, includes some incredible underwater field work in Lake Nicaragua and its satellite crater lakes. But George is perhaps best known for his laboratory work on mate choice and aggression in these same fish. I reviewed that body of work in some older TFH columns (See “Wayne’s New World: ACA Honors Dr. George W. Barlow, Parts 1 and 2,” TFH March and April 1996) but George does a much better job of it in his recent, popular book, The Cichlid Fishes: Nature’s Grand Experiment in Evolution (Perseus Publications, 2000), a must-read for all cichlidophiles. Check out his chapters titled “Oh Yeah? Put Up Your Fins!” and “Mating Gets Personal.” Or better, read the entire book! In a nutshell, his work with Midas cichlids, a typical monogamous biparental substrate-brooder in which males hold territories and both parents work to protect and rear their young, shows that females are choosy, and they choose males who demonstrate vigorous aggressiveness and the ability to hold and defend choice territories.

Parental Investment

Since the reproductive success of monogamous biparental cichlids like the Midas cichlid requires substantial parental investment from both parents, both sexes ought to be selective when mating. When offered a choice (in a two-way or three-way choice apparatus), restrained females (behind a clear barrier) preferentially select the largest available male and, independently, the most aggressive male. These two male characteristics, size and aggressiveness, are often related, but not always.

Paired Battles

When males are fought in pair-wise fashion, typically the larger of the two wins the encounter and dominates the loser. In the case of the rare gold morph of the Midas cichlid, which aquarists prefer (versus the gray, barred common morph), males that are smaller than their wild, gray counterparts typically win these paired battles. Something about the rare gold coloration adds apparent dominance to these gold fish. Paradoxically, the female’s choice in an experimental but artificial “choose-a-tron” apparatus (as just described) is not always the best choice for her.

Female Selection

In half the cases in which females preferentially select the larger or more aggressive male, when the barrier is removed that male aggressively rejects her! Thus, males are choosy too, and mate choice depends on compatibility, a female that gives as good as she gets. In the wild, females attempt to select and attract big, aggressive males that can hold and defend territory, and males screen potential mates for their own aggressiveness and size (big females lay more eggs and can help with defense). Also, controlling for size (males and females of identical size), it is males that are inherently more aggressive than females.


As we cichlidophiles all know (and enjoy watching, for the most part, as long as it succeeds!), pair formation is characterized by prolonged and complex ritualized behavioral exchanges between the male and female. There are the usual gill flaring, frontal approach, fin erection, lateral displaying, and even jaw-locking/wrestling behaviors that signal interest between the sexes. These combative behaviors during pair formation are believed to be a means of testing one another’s suitability as a mate, and they also establish the dominance relationship between male and female.

Larger Males

Aggression and the consequent establishment of dominance relationships is thus the central feature of pair bonding. If the two fish succeed in pairing, they then direct their aggression outward against conspecifics and other fishes, a feature exploited by cichlid hobbyists in the form of additional target fish, which are often added to cement shaky pair bonds. When pairing fails in the wild, the loser is vanquished and swims away to find a more compatible, less aggressive mate. In the relatively small confines of the typical aquarium, the rejected suitor, typically the female, tries to flee and is attacked by the rejecting male until death or removal results. These are the behaviors that we cichlidophiles who have spawned our fish have witnessed empirically in our tanks.

Barlow spent the greater part of his career quantifying in controlled experiments how this all works in the Midas cichlid as a model system for all cichlids—at least other monogamous biparental substrate-brooders. And, in those other cichlid species that have received experimental scrutiny (e.g., convict cichlids for one), these generalizations about mate choice and pair formation seem to (pardon me) hold water. Males tend to be bigger than females both at full-grown size and in naturally formed pairs, and both sexes are choosy when it comes to selecting mates, basing their choices on relative size and on aggressiveness as indicators of the fitness of a potential partner in cooperatively raising young.

Larger Females

But often exceptions are more interesting than the rule. One exception that recently fell into the Barlow behavioral viewfinder was species of the genus Julidochromis from Lake Tanganyika. Again, most cichlidophiles are familiar with these wonderful aquarium fish known commonly as “julies.” Even I, a diehard New World fan, have kept and raised them, and for many years at that! If you have kept them too you will probably recall that females of all known Julidochromis species are typically much larger than their male consorts in a breeding pair. This is an obvious reversal in these monogamous, substrate-brooders (but philopatric: other related youngsters help in the rearing of successive clutches, and the substrate is the interstices of a rock pile) and an exception for pair-bonded cichlid fish. This apparently is also true for Julidochromis species in the wild.

Dominant Females

Yamagishi and Kohda (1996, “Is the Cichlid Fish Julidochromis marlieri Polyandrous?” Ichthyol. Res. 43: 469-471) reported that females of bonded pairs of J. marlieri in Lake Tanganyika are normally larger than their mates and appear to be the dominant partner. Although size distributions of males and females overlap widely, paired males were consistently only about 75 percent of the length of their partners. Barlow and Lee (2002) estimated from the length/weight regression of captive J. marlieri that the typical paired male is thus, on average, 56 percent of the mass of his mate! Other than size, males and females are so alike in appearance that unpaired fish have to be sexed via examination of their spawning papilla. Also, both male and female parents participate equally in all aspects of egg, larvae, and fry defense and rearing: there is no clear-cut separation of duty as is true for many other monogamous biparental cichlids.

Females usually defend territories containing a single male in the wild, but in one instance it’s a female paired simultaneously with two (Yamagishi and Kohda, 1996). So, perhaps given the reverse size differential, larger female Julidochromis might be facultatively polyandrous, pairing with multiple males simultaneously when the opportunity was right (as opposed to facultative, opportunistic polygyny, which is often seen in typical “monogamous” cichlids). Moreover, this size differential might reflect the evolutionary result of male rarity in nature. Most interesting.

Size and Dominance

Barlow wondered whether this size reversal in paired julies reflected a general dominance of females over males in these species. If so, was it size alone that accounted for female dominance over males, or was aggressiveness in these fish also reversed? The experiments that Barlow and Lee devised and ran were designed to answer these questions. They placed individual males and females together in a situation that they knew from experience would elicit immediate fighting. They varied the relative sizes of male and female combatants to determine the relationship between size and the probability of winning a fight. If one sex were more likely to win when the two were equal in size, that result would indicate an inherent difference in aggressiveness; Barlow and Lee predicted that females would win.


Beyond that, they also wished to determine the relative size at which the male and female would have equal chances of winning—they predicted that such a male would have to be larger than the female. To measure “dominance” in these combats, they determined which fish initiated the fight and, once a fight had started, which fish escalated the fight fastest. They expected that females would initiate more and escalate faster compared with males, even at identical size.

Sample Size

Their subjects were 17 male and 22 female adult J. marlieri (sexed by genital examination) that were wild-caught from Burundi along the shore of Lake Tanganyika. These were housed individually until tested. A total of 29 fights were conducted variously between this group, varying the size relationship of the two combatants, and though a few were reused, no two fish were ever rematched.


The fights were initiated simply by placing the two test fish on either side of an opaque divider in a 30-gallon long tank, allowing them to acclimate for 1½ hours at which point the divider was removed. They noted which fish was the first to do any of the following: approach, roll (on its long axis directing its dorsal fin at the other fish), spread its medial fins, tail beat via body undulation, or bite following an accelerating dash. The winner was the first fish from which the other fled three times in quick succession at which time the loser was removed. The time that elapsed between removal of the barrier and declaration of a winner was also noted. A typical julie fight would start with an approach by one or both fish followed by rolling and fin erection. This was followed by tail beating escalating to biting and finally mouth pushing in which the fish face one another, make contact with their open mouths, and push. This, in turn, often escalates to mouth locking and finally release and fleeing.

Fascinating Data

Their findings: the sexes did not differ significantly in time taken to initiate fights, but females were significantly more inclined to escalate than were males. In that sense they were the more aggressive sex. Females consistently won fights when they were the same size or bigger than males. At 90 percent the mass (size to weight ratio) of a male, males have an equal chance of winning. Since paired males average only about 56 percent of the mass of their mates in nature, females should easily dominate their mates, since they are, on average, almost twice as heavy.

One might expect the optimum compatibility match would be between a male and female who is about 90 percent of his size; this pair would have equally dominant mates. But that’s not what either sex selects in nature where seemingly endless choices exist. The fact that this doesn’t happen, Barlow and Lee argue, suggests that in the case of J. marlieri there is selective advantage in having one individual—in this case the female—constantly in charge, whose dominance in the “relationship” is never in question (the “heavy” of the pair, so to speak!). And if the female is the dominant sex in pairs, this would allow mating with multiple males to occur (polyandry), which translates into a higher potential rate of reproduction for the female, just the reverse of the situation for most monogamous biparental cichlids whose males will opportunistically mate with additional females. How and why did this arise? Clearly, female Julidochromis must be competing for access to a limited number of males in Lake Tanganyika, suggests Barlow.

As with most of Barlow’s experimental behavioral work, this small study is an exemplar of important questions framed simply and whose answers are eminently testable in an equally simple experimental paradigm. It also represents an intriguing initial contribution to understanding the increasingly exceptional biology of a group of unique, polyandrous cichlids.

Looking Ahead

Though George Barlow is no longer here to finish the work he started just a few short years ago with Julidochromis, others are and have been doing so, standing on the shoulders of this early pioneer of cichlid behavior. In a future column I will continue our discussion of this unique suite of cichlids based on more recently published work, both in the field and in the laboratory, that I discovered while writing this piece. Meanwhile, as Ron Coleman, a dear friend of and former post-doctoral fellow in Barlow’s lab suggested to us at ACA 2007 in tribute to George: take the time to sit and really watch your fish. That’s what George Barlow really was about, understanding what cichlids do and why. And he counted on and encouraged us aquarists to generate those interesting new observations that would provide leads for future cichlid research. So let’s do it—in his memory.

See the full article on TFH Digital