Cleaner fish recognize self in a mirror via self-face recognition like humans.

Some animals have the remarkable capacity for mirror self-recognition (MSR), yet any implications for self-awareness remain uncertain and controversial. This is largely because explicit tests of the two potential mechanisms underlying MSR are still lacking: mental image of the self and kinesthetic visual matching. Here, we test the hypothesis that MSR ability in cleaner fish, Labroides dimidiatus, is associated with a mental image of the self, in particular the self-face, like in humans. Mirror-naive fish initially attacked photograph models of both themselves and unfamiliar strangers. In contrast, after all fish had passed the mirror mark test, fish did not attack their own (motionless) images, but still frequently attacked those of unfamiliar individuals. When fish were exposed to composite photographs, the self-face/unfamiliar body were not attacked, but photographs of unfamiliar face/self-body were attacked, demonstrating that cleaner fish with MSR capacity recognize their own facial characteristics in photographs. Additionally, when presented with self-photographs with a mark placed on the throat, unmarked mirror-experienced cleaner fish demonstrated throat-scraping behaviors. When combined, our results provide clear evidence that cleaner fish recognize themselves in photographs and that the likely mechanism for MSR is associated with a mental image of the self-face, not a kinesthetic visual-matching model. Humans are also capable of having a mental image of the self-face, which is considered an example of private self-awareness. We demonstrate that combining mirror test experiments with photographs has enormous potential to further our understanding of the evolution of cognitive processes and private self-awareness across nonhuman animals.

Further evidence for the capacity of mirror self-recognition in cleaner fish and the significance of ecologically relevant marks.

An animal that tries to remove a mark from its body that is only visible when looking into a mirror displays the capacity for mirror self-recognition (MSR), which has been interpreted as evidence for self-awareness. Conservative interpretations of existing data conclude that convincing evidence for MSR is currently restricted to great apes. Here, we address proposed shortcomings of a previous study on MSR in the cleaner wrasse Labroides dimidiatus, by varying preexposure to mirrors and by marking individuals with different colors. We found that (1) 14/14 new individuals scraped their throat when a brown mark had been provisioned, but only in the presence of a mirror; (2) blue and green color marks did not elicit scraping; (3) intentionally injecting the mark deeper beneath the skin reliably elicited spontaneous scraping in the absence of a mirror; (4) mirror-naive individuals injected with a brown mark scraped their throat with lower probability and/or lower frequency compared to mirror-experienced individuals; (5) in contrast to the mirror images, seeing another fish with the same marking did not induce throat scraping; and (6) moving the mirror to another location did not elicit renewed aggression in mirror-experienced individuals. Taken together, these results increase our confidence that cleaner fish indeed pass the mark test, although only if it is presented in ecologically relevant contexts. Therefore, we reiterate the conclusion of the previous study that either self-awareness in animals or the validity of the mirror test needs to be revised.

Male Guppies Recognize Familiar Conspecific Males by Their Face.

Individual recognition is a necessary cognitive ability for the maintenance of stable social relationships. Recent studies have shown that like primates, some fish species can distinguish familiar fish from unfamiliar strangers via face-recognition. However, the taxa of the studied fish species are restricted (within Perciformes) and the visual signal used for the recognition of fish remains unclear. Here, we investigated the visual signal for individual-recognition in males of a sexually dichromatic guppy (Poecilia reticulata, Cyprinodontiformes). Using guppy males, we examined the hypothesis that fish distinguish between familiar individuals and unknown strangers by their faces rather than by body coloration. We randomly presented focal fish with four types of composite photo-models: familiar (familiar-face and familiar-body = F/F), stranger (stranger-face and stranger-body = S/S), familiar face combined with stranger body (F/S) and stranger face combined with familiar body (S/F). Focal males infrequently attacked familiar-face models but frequently attacked stranger-face models, regardless of body types. These behavioral reactions indicate that guppy males discriminate between familiar and stranger males by their face, not body coloration with wide variation. Importantly, male faces contain clear individual-variation in white/metallic colored patches on the operculum visible for humans. Considering the photo-model, our results suggest that these patches might be an important visual stimulus for face-recognition in guppy males, like some cichlids. Comparative examination among males of different guppy variants, including wild type phenotype, suggests that the face color-patch is stable regardless of variation in body color, with a different genetic mechanism potentially underlying face and body colors.

If a fish can pass the mark test, what are the implications for consciousness and self-awareness testing in animals?

Abstract: The ability to perceive and recognise a reflected mirror image as self (mirror self-recognition, MSR) is considered a hallmark of cognition across species. Although MSR has been reported in mammals and birds, it is not known to occur in any other major taxon. Potentially limiting our ability to test for MSR in other taxa is that the established assay, the mark test, requires that animals display contingency testing and self-directed behaviour. These behaviours may be difficult for humans to interpret in taxonomically divergent animals, especially those that lack the dexterity (or limbs) required to touch a mark. Here, we show that a fish, the cleaner wrasse Labroides dimidiatus, shows behaviour that may reasonably be interpreted as passing through all phases of the mark test: (i) social reactions towards the reflection, (ii) repeated idiosyncratic behaviours towards the mirror, and (iii) frequent observation of their reflection. When subsequently provided with a coloured tag in a modified mark test, fish attempt to remove the mark by scraping their body in the presence of a mirror but show no response towards transparent marks or to coloured marks in the absence of a mirror. This remarkable finding presents a challenge to our interpretation of the mark test­do we accept that these behavioural responses, which are taken as evidence of self-recognition in other species during the mark test, lead to the conclusion that fish are self-aware? Or do we rather decide that these behavioural patterns have a basis in a cognitive process other than self-recognition and that fish do not pass the mark test? If the former, what does this mean for our understanding of animal intelligence? If the latter, what does this mean for our application and interpretation of the mark test as a metric for animal cognitive abilities? EDITOR'S NOTE: This Short Report received both positive and negative reviews by experts. The Academic Editor has written an accompanying Primer that we are publishing alongside this article (https://doi.org/10.1371/journal.pbio.3000112). The linked Primer presents a complementary expert perspective; it discusses how the current study should be interpreted in the context of evidence for and against self-awareness in a wide range of animals.

Sperm Allocation Strategies Depending on Female Quality in Medaka (Oryzias latipes).

Sperm production is costly, and males are expected to strategically allocate this potentially limiting resource to maximize their fitness. Sperm allocation theory predicts that males should adjust their sperm expenditure in relation to female quality. However, the available empirical evidence is limited. In this study, we assessed whether wild male medaka (Oryzias latipes) would allocate their sperm depending on female quality under controlled conditions. Behavioral observations revealed that spawning behavior could be classified into four stages: male quivering, female quivering, sperm release, and stay. Of these behaviors, only the duration of sperm release was positively correlated with the number of sperm ejaculated in the aquarium, suggesting that males may adjust the sperm number expended for each mating by controlling the duration of sperm release. The estimated amount of sperm released per spawning was positively and significantly correlated with the body depth and weight of paired females. This result is consistent with the prediction arising from the sperm allocation hypothesis, and suggests that male medaka may allocate their sperm depending on paired female body depth and weight. However, the number of released sperm did not correlate with the number of spawned eggs, which was positively correlated with body depth and weight of females. These results imply that medaka females may have counter-strategies against sperm allocation by males, and consequently, males might not be able to adjust the number of sperm efficiently. We disscuss our results from the perspective of sexual conflict over sperm as a limited resource.

Does a cichlid fish process face holistically? Evidence of the face inversion effect.

Faces are the most important body part for differentiating among human individuals by humans. Humans read the face as a whole, rather than looking at its parts, which makes it more difficult to recognise inverted faces than upright. Some other mammals also identify each other based on the upright face and take longer to recognise inverted faces. This effect is called the face inversion effect and is considered as evidence for face-specific perception. This ability has rarely been observed in animals other than mammals, but it was recently reported that some fish species could distinguish among individuals based on the face. For example, the cichlid fish Neolamprologus pulcher rapidly recognises familiar conspecifics by faces rather than other body parts. Here, we examined the face inversion effect in N. pulcher, by showing photographs of conspecific fish faces and objects in both upright and inverted orientations. Subjects gazed at novel faces longer than familiar faces in upright presentation, whereas they did not show such a tendency for inverted faces. Although the object discrimination was difficult, we did not observe the difference between upright and inverted object photographs. Our results indicate that fish exhibits the inversion effect for faces. These findings suggest that N. pulcher may process their conspecifics' face holistically, like humans.

Do Scales of the Cichlid Altolamprologus compressiceps in Lake Tanganyika Function as a Morphological Defense Against Scale-Eating?

Many cichlid species in the shallow-shore of Lake Tanganyika suffer damage from attacks by the scale-eater Perissodus microlepis. Many prey fish engage in warning behaviors to this predator. It has been hypothesized that, if prey fish have difficulty employing such behavioral tactics, morphological defenses against scale-eating, such as hard scales, will evolve. The shrimp-eating cichlids, Altolamprologus compressiceps (Ac) and Neolamprologus fasciatus (Nf), exhibit hunting behaviors in which they remain motionless for up to 10 seconds while aiming at prey, when they are vulnerable to scale-eating predators; thus, these fish have likely evolved morphological defenses against "scale-attacks". We tested this hypothesis in Ac and Nf, as well as three other predatory fish, Lamprologus callipterus, Lepidiolamprologus elongatus and Lep. attenuatus, that are not motionless for such a long time. Under natural conditions, Ac and Nf were rarely attacked, while the other three species were attacked frequently. When freshly killed specimens of these five species were displayed underwater in the presence of P. microlepis, Ac was rarely attacked, while Nf and the three other species were attacked frequently. Among the five fish species, the force required to tear off scales was highest for Ac, and this force was negatively correlated with the frequency of attacks on the displayed fish. These results support the hypothesis that the hard scales of Ac function as an anti-scale-attack measure, although it remains unclear why free-swimming Nf were rarely attacked while aiming at prey, despite the fact that the force required to tear off its scales was not large.

Variations in the breeding behavior of cichlids and the evolution of the multi-functional seminal plasma protein, seminal plasma glycoprotein 120.

BACKGROUND: Seminal plasma proteins are associated with successful fertilization. However, their evolutionary correlation with fertilization mechanisms remains unclear. Cichlids from Lake Tanganyika show a variety-rich spawning behavior that is associated with the transfer of the sperm to the egg for fertilization. One of these behaviors, called "oral fertilization," emerged during their speciation. In oral fertilization, females nuzzle the milt from male genitalia and pick up the released eggs in their mouths, which are then fertilized inside the oral cavity. Thus, the success of the fertilization is dependent on the retention of sperm in the oral cavity during spawning. Sperm aggregation and immobilization in viscous seminal plasma may help retain the sperm inside the oral cavity, which ultimately determines the success of the fertilization. Seminal plasma glycoprotein 120 (SPP120) is one of the major seminal plasma proteins present in cichlids. SPP120 has been implicated to immobilize sperm and increase the milt viscosity. However, the functional linkage between oral fertilization and seminal plasma proteins has not been investigated. RESULTS: During trials of simulated oral fertilization, it was observed that milt viscosity contributed to fertilization success by facilitating longer retention of the milt inside the mouth during spawning. Glycosylation of SPP120 was associated with high milt viscosity. Its glycosylation was specifically present in the milt of cichlid species exhibiting oral fertilization. Moreover, recombinant SPP120 from several the oral fertilization species strongly immobilized/aggregated sperm. Therefore, the functions of SPP120 (immobilization/aggregation and its glycosylation) may contribute to success of oral fertilization, and these functions of SPP120 are more prominent in oral fertilization species. In addition, comparative phylogenetic analyses showed a positive evolutionary correlation between SPP120 function and oral fertilization. Hence, these evolutions may have occurred to keep up with the transition in the mode of fertilization. In addition, rapid evolution in the molecular sequence might be associated with functional modifications of SPP120. CONCLUSION: These results suggest that SPP120 might be associated with oral fertilization. In other words, reproductive traits that define the interaction between sperms and eggs could be the evolutionary selective force that cause the rapid functional modification of the fertilization-related reproductive protein, SPP120.

A social cichlid fish failed to pass the mark test.

Since the pioneering work in chimpanzees, mirror self-recognition (MSR), the ability to recognise oneself in a mirror, has been reported in great apes, Asian elephants, dolphins, and some social birds using the mark test, in which animals that possess MSR touch an imperceptible mark on their own bodies only when a mirror is present. However, giant pandas, which are solitary, failed to pass the mark test, suggesting that MSR evolved solely in highly social animals. In contrast to the increasing evidence of MSR in mammals and birds, little is known about MSR in fish. A Tanganyikan cichlid, Neolamprologus pulcher, is a good candidate for study because these fish live in highly social groups and recognise conspecifics about as rapidly as primates. We examined their responses to a mirror image and tested whether N. pulcher could pass the mark test. When the mirror was first exposed, they stayed in front of the mirror and exhibited aggressive behaviour towards the mirror image. These social behaviours suggested that the focal fish perceived the mirror image as an unfamiliar conspecific. The social responses decreased over the following days, as has generally been the case in animals with MSR. After mark injection, we found no increase in scraping behaviour or prolonged observation of the marked side. These results show a lack of contingency checking and mark-directed behaviours, meaning that N. pulcher failed to pass the mark test and did not recognise their self-image in the mirror.

Laterality is Universal Among Fishes but Increasingly Cryptic Among Derived Groups.

Laterality has been studied in several vertebrates, mainly in terms of brain lateralization and behavioral laterality, but morphological asymmetry has not been extensively investigated. Asymmetry in fishes was first described in scale-eating cichlids from Lake Tanganyika, in the form of bilateral dimorphism in which some individuals, when opening their mouths, twist them to the right and others to the left. This asymmetry has a genetic basis, and is correlated with lateralized attack behaviors. This has subsequently been found in fishes from numerous taxa with various feeding habits. The generality of such morphological laterality should thus be investigated in as wide a range of fishes as possible. Using specific indicators of lateral differences in mandibles and head inclination, we find that representative species from all 60 orders of extant gnathostome fishes (both bony and cartilaginous) possess morphological laterality. Furthermore, we identify the same laterality in agnathans (hagfish and lamprey), suggesting that this trait appeared early in fish evolution and has been maintained across fish lineages. However, a comparison of asymmetry among groups of bony fishes reveals, unexpectedly, that phylogenetically more recent-groups possess less asymmetry in body structures. The universality of laterality in fishes indicates a monophyletic origin, and may have been present in the ancestors of vertebrates. Ecological factors, predator-prey interactions in particular, may be key drivers in the evolution and maintenance of dimorphism, and may also be responsible for the cryptic trend of asymmetry in derived groups. Because lungfish and coelacanths share this trait, it is likely that tetrapods also inherited it. We believe that study of this morphological laterality will provide insights into the behavioral and sensory lateralization of vertebrates.

The social and ecological costs of an 'over-extended' phenotype.

Extended phenotypes offer a unique opportunity to experimentally manipulate and identify sources of selection acting on traits under natural conditions. The social cichlid fish Neolamprologus multifasciatus builds nests by digging up aquatic snail shells, creating an extended sexual phenotype that is highly amenable to experimental manipulation through addition of extra shells. Here, we find sources of both positive sexual selection and opposing natural selection acting on this trait; augmenting shell nests increases access to mates, but also increases social aggression and predation risk. Increasing the attractiveness of one male also changed social interactions throughout the social network and altered the entire community structure. Manipulated males produced and received more displays from neighbouring females, who also joined augmented male territories at higher rates than unmanipulated groups. However, males in more attractive territories received more aggression from neighbouring males, potentially as a form of social policing. We also detected a significant ecological cost of the 'over-extended' phenotype; heterospecific predators usurped augmented nests at higher rates, using them as breeding sites and displacing residents. Using these natural experiments, we find that both social and ecological interactions generate clear sources of selection mediating the expression of an extended phenotype in the wild.

Diet disparity among sympatric herbivorous cichlids in the same ecomorphs in Lake Tanganyika: amplicon pyrosequences on algal farms and stomach contents.

BACKGROUND: Lake Tanganyika, an ancient lake in the Great Rift Valley, is famous for the adaptive radiation of cichlids. Five tribes of the Cichlidae family have acquired herbivory, with five ecomorphs: grazers, browsers, scrapers, biters and scoopers. Sixteen species of the herbivorous cichlids coexist on a rocky littoral slope in the lake. Seven of them individually defend feeding territories against intruding herbivores to establish algal farms. We collected epiphyton from these territories at various depths and also gathered fish specimens. Algal and cyanobacteria community structures were analysed using the amplicon-metagenomic method. RESULTS: Based on 454-pyrosequencing of SSU rRNA gene sequences, we identified 300 phototrophic taxa, including 197 cyanobacteria, 57 bacillariophytes, and 31 chlorophytes. Algal farms differed significantly in their composition among cichlid species, even in the same ecomorph, due in part to their habitat-depth segregation. The algal species composition of the stomach contents and algal farms of each species differed, suggesting that cichlids selectively harvest their farms. The stomach contents were highly diverse, even between species in the same tribe, in the same feeding ecomorph. CONCLUSIONS: In this study, the amplicon-metagenomic approach revealed food niche separation based on habitat-depth segregation among coexisting herbivorous cichlids in the same ecomorphs in Lake Tanganyika.

Duration of memory of dominance relationships in a group living cichlid.

Animal contests are costly and tend to escalate when rivals have similar competitive abilities. Individuals that remember dominance relationships with rivals may avoid repeated agonistic interactions and hence avoid the costs of repeated escalation of contests. However, it can be difficult to experimentally disentangle the effects of memory from those of loser effects (losers behaving subordinately due to prior defeats). Here, we test whether loser effects or individual memory mediate contest behaviour in the African cichlid, Julidochromis transcriptus. We find that on days 3 and 5 after initial contests, losers display subordinate behaviour to contest winners but not to novel contestants. However, this effect disappears after 7 days, at which time losers do not display subordinate behaviour to either rival. These results show that (1) this fish can recall a previously dominant contestant for up to 5 days and (2) as no subordinate displays were shown to the novel contestant, there are no evidences for loser effects in this species. Such short-term memory of past interactions may have broad significance in social species with repeated interactions.

Territorial males can sire more offspring in nests with smaller doors in the cichlid Lamprologus lemairii.

To examine how territorial males counter reproductive parasites, we examined the paternity of broods guarded by territorial males using 5 microsatellite loci and factors that determine siring success in a wild population of the Lake Tanganyika cichlid Lamprologus lemairii. Females enter rock holes (nests) and spawn inside, and territorial males release milt over the nest openings. Sneakers attempt to dart into the nests, but territorial males often interrupt the attempt. The body size of territorial males (territorial defense ability) and the size of nest opening (the ability to prevent sneakers from nest intrusions) are predicted to be factors that affect paternity at the premating stage, whereas milt quality traits are factors that affect paternity at the postmating stage. Parentage analyses of 477 offspring revealed that most clutches have few or no cuckolders, and territorial males sired >80% of eggs in 7 of the 10 analyzed clutches. Larger territorial males that spawned in nests with narrower openings had greater siring success. In contrast, none of the milt traits affected the siring success. These suggest that territorial male L. lemairii adopt premating strategies whereby they effectively prevent reproductive parasitism.

Cleaner fish with mirror self-recognition capacity precisely realize their body size based on their mental image.

Animals exhibiting mirror self-recognition (MSR) are considered self-aware; however, studies on their level of self-awareness remain inconclusive. Recent research has indicated the potential for cleaner fish (Labroides dimidiatus) to possess a sophisticated level of private self-awareness. However, as this study revealed only an aspect of private self-awareness, further investigation into other elements is essential to substantiate this hypothesis. Here, we show that cleaner fish, having attained MSR, construct a mental image of their bodies by investigating their ability to recall body size. A size-based hierarchy governs the outcomes of their confrontations. The mirror-naïve fish behaved aggressively when presented with photographs of two unfamiliar conspecifics that were 10% larger and 10% smaller than their body sizes. After passing the MSR test, they refrained from aggression toward the larger photographs but still behaved aggressively toward the smaller ones without re-examining their mirror images. These findings suggest that cleaner fish accurately recognize their body size based on mental images of their bodies formed through MSR. Additionally, mirror-experienced fish frequently revisited the mirror when presented with an intimidating larger photograph, implying the potential use of mirrors for assessing body size. Our study established cleaner fish as the first non-human animal to be demonstrated to possess private self-awareness.

Invisible pair bonds detected by molecular analyses.

A focus on pair bonds between males and females is fundamental to study the evolution of social organization. Because pair bonds are generally identified from direct observations of pairs that maintain physical proximity, pair bonds may have been overlooked in animals that do not exhibit such visible pairs. The Lake Tanganyika cichlid fish Xenotilapia rotundiventralis forms schools that consist of mouthbrooding and non-brooding adults in mid-water, and visible pairs are not recognized. A previous study suggested that mouthbrooding females transfer fractions of the young to males when the young become large. However, it remains a mystery whether the mating pairs maintain pair bonds so that the females can transfer the young to their mates. To answer this question, we conducted a parentage analysis using 10 microsatellite markers. The analysis showed that the mouthbrooding adults were most likely genetic fathers and mothers of the young in their mouths. This finding suggests that the female-to-male shift of young takes place between mating partners, and thus the mating pairs maintain pair bonds at least until the shift of young. The present study is the first to detect pair bonds in animals in which physical proximity has not been observed.

Changes in reproductive life-history strategies in response to nest density in a shell-brooding cichlid, Telmatochromis vittatus.

To determine whether the appearance of a reproductively parasitic tactic varies, and how this variation affects territorial males of the Lake Tanganyika cichlid fish Telmatochromis vittatus, we examined the reproductive ecology of territorial males in Mtondwe and compared it with that of a neighboring Wonzye population, where nest density differs from that at Mtondwe. In Wonzye, with high nest density, male tactics change with their body size from a territorial to a non-territorial parasitic tactic called piracy in which they conquer several nests defended by territorial males and take over the nests while females are spawning. These "pirate" males could decrease the costs incurred by travelling among nests by exclusively targeting aggregations of nests in close proximity while avoiding separate nests. Territorial males in Wonzye sacrifice the potential higher attractiveness offered by large nests and instead compete for nests farther from neighbors on which pirates less frequently intrude. In contrast, the Mtondwe population had lower nest density and piracy was absent. Given that the success of piracy depends on the close proximity of nests, nest density is likely responsible for the observed variation in the occurrence of piracy between the two populations. Furthermore, in Mtondwe, territorial males competed for larger nests and were smaller than the territorial males in Wonzye. Thus, this lower nest density may free territorial males from the selection pressures for increased size caused by both defense against nest piracy and the need to develop into pirates as they grow.

Inheritance patterns of lateral dimorphism examined through breeding experiments in Tanganyikan cichlid (Julidochromis transcriptus) and Japanese Medaka (Oryzias latipes).

The asymmetry of mouth morphology, in which the mouth opens either leftward or rightward, is a unique example of antisymmetry that is widely found in fishes, such as Tanganyikan scale-eating cichlids, herbivorous cichlids, and the Japanese freshwater goby. This dimorphism is thought to be heritable in a Mendelian manner similar to that of the dominance of the lefty allele over righty, with homozygotes of lefty alleles being absent. This study aims to reveal whether this trait is inherited in the same Mendelian manner in fishes other than those examined to date, and whether the absence of the dominant homozygote is due to a lethal effect. We conducted F(1) and F(2) breeding experiments using the Tanganyikan cichlid Julidochromis transcriptus and Japanese medaka Oryzias latipes. For both J. transcriptus and O. latipes, the F(1) generations produced by incrossing lefty parents showed a lefty:righty ratio of 2:1, whereas incrosses of righties produced only righty offspring. Test crosses between lefty and righty produced a 1:1 ratio of lefty and righty offspring. These results are consistent with the inheritance model in which righty is homozygous for a recessive righty allele and lefty is heterozygous for dominant lefty and recessive righty alleles, in agreement with previous observations. The F(2) test cross also confirmed this pattern. Furthermore, no lethal effects on hatchability of O. latipes were detected, refuting the suggestion that the dominant homozygote is embryonic lethal. We conclude that mouth laterality is inherited in the same Mendelian manner in these two species, similar to that other fish species studied previously. However, the reason for absence of the dominant homozygote remains unclear.

Parker's sneak-guard model revisited: why do reproductively parasitic males heavily invest in testes?

Alternative reproductive tactics are widespread in males and may cause intraspecific differences in testes investment. Parker's sneak-guard model predicts that sneaker males, who mate under sperm competition risk, invest in testes relatively more than bourgeois conspecifics that have lower risk. Given that sneakers are much smaller than bourgeois males, sneakers may increase testes investment to overcome their limited sperm productivity because of their small body sizes. In this study, we examined the mechanism that mediates differential testes investment across tactics in the Lake Tanganyika cichlid fish Lamprologus callipterus. In the Rumonge population of Burundi, bourgeois males are small compared with those in other populations and have a body size close to sneaky dwarf males. Therefore, if differences in relative testis investment depend on sperm competition, the rank order of relative testis investment should be dwarf males > bourgeois males in Rumonge = bourgeois males in the other populations. If differences in relative testis investment depend on body size, the rank order of relative testes investment should be dwarf males > bourgeois males in Rumonge > bourgeois males in the other populations. Comparisons of relative testis investment among the three male groups supported the role of sperm competition, as predicted by the sneak-guard model. Nevertheless, the effects of absolute body size on testes investment should be considered to understand the mechanisms underlying intraspecific variation in testes investment caused by alternative reproductive tactics.