Filial cannibalism in fishes: Why do parents eat their offspring?

Filial cannibalism (the eating of one's own offspring) occurs in a variety of taxa, but is especially prevalent in fishes with parental care. Recent research supports a central tenet of parental-investment theory; that is, parents consume their offspring when it maximizes their lifetime reproductive success. This review outlines the theoretical framework used to explain the adaptive significance of filial cannibalism, evaluates experimental studies to test some predictions of this theory and discusses how the occurrence of filial cannibalism affects other aspects of a species' reproductive ecology.     

Parents benefit from eating offspring: density-dependent egg survivorship compensates for filial cannibalism

Abstract Why should animals knowingly consume their own young? It is difficult to imagine many circumstances in which eating one's own young (i.e., filial cannibalism) actually increases an individual's fitness; however, filial cannibalism commonly co‐occurs with parental care in fishes. The evolutionary significance of filial cannibalism remains unclear. The most commonly accepted explanation is that filial cannibalism is a mechanism by which caring males gain energy or nutrients that they reinvest into future reproduction, thereby increasing net reproductive success. There is mixed support for this hypothesis and, at best, it can only explain filial cannibalism in some species. A recent alternative hypothesis suggests that filial cannibalism improves the survivorship of remaining eggs by increasing oxygen availability, and thus increases current reproductive success. This theory has received little attention as of yet. We evaluated the hypothesis of oxygen‐mediated filial cannibalism in the sand goby by examining the effect of oxygen and egg density on the occurrence of filial cannibalism, evaluating the effects of partial clutch cannibalism on the survivorship of remaining eggs, and comparing potential costs and benefits of filial cannibalism related to the net number of eggs surviving. Indeed, we found that oxygen level and egg density affected the occurrence of cannibalism and that simulated partial clutch cannibalism improved survivorship of the remaining eggs. Additionally, because increased egg survivorship, stemming from partial egg removal, compensated for the cost of cannibalism (i.e., number of eggs removed) at a range of cannibalism levels, filial cannibalism potentially results in no net losses in reproductive success. However, oxygen did not affect egg survivorship. Thus, we suggest a more general hypothesis of filial cannibalism mediated by density‐dependent egg survivorship.     

Effects of among-offspring relatedness on the origins and evolution of parental care and filial cannibalism

Parental care is expected to increase the likelihood of offspring survival at the cost of investment in future reproductive success. However, alternative parental behaviours, such as filial cannibalism, can decrease current reproductive success and consequently individual fitness. We evaluate the role of among-offspring relatedness on the evolution of parental care and filial cannibalism. Building on our previous work, we show how the evolution of care is influenced by the effect of among-offspring relatedness on both the strength of competition and filial cannibalism. When there is a positive relationship between among-offspring competition and relatedness, parental care will be favoured when among-offspring relatedness is relatively low, and the maintenance of both care and no-care strategies is expected. If the relationship between among-offspring competition and relatedness is negative, parental care is most strongly favoured when broods contain highly related offspring. Further, we highlight the range of conditions over which the level of this among-offspring relatedness can affect the co-occurrence of different care/no care and cannibalism/no cannibalism strategies. Coexistence of multiple strategies is independent of the effects of among-offspring relatedness on cannibalism but more likely when among-offspring relatedness and competition are positively associated.     

When to care for, abandon, or eat your offspring: the evolution of parental care and filial cannibalism

Parental care and filial cannibalism (the consumption of one's own offspring) co-occur in many animals. While parental care typically increases offspring survival, filial cannibalism involves the killing of one's young. Using an evolutionary ecology approach, we evaluate the importance of a range of factors on the evolution of parental care and filial cannibalism. Parental care, no care/total abandonment, and filial cannibalism evolved and often coexisted over a range of parameter space. While no single benefit was essential for the evolution of filial cannibalism, benefits associated with adult or offspring survival and/or reproduction facilitated the evolution of cannibalism. Our model highlights the plausibility of a range of alternative hypotheses. Specifically, the evolution of filial cannibalism was enhanced if (1) parents could selectively cannibalize lower-quality offspring, (2) filial cannibalism increased egg maturation rate, (3) energetic benefits of eggs existed, or (4) cannibalism increased a parent's reproductive rate (e.g., through mate attractiveness). Density-dependent egg survivorship alone did not favor the evolution of cannibalism. However, when egg survival was density dependent, filial cannibalism invaded more often when the density dependence was relatively more intense. Our results suggest that population-level resource competition potentially plays an important role in the evolution of both parental care and filial cannibalism.     

Filial Cannibalism in the Biparental Fish Cichlasoma nigrofasciatum (Pisces: Cichlidae) in Response to Early Brood Reductions

Filial cannibalism (eating one's own offspring) may enhance a parent's lifetime reproductive success if the costs associated with this behaviour are outweighed by its benefits. An organism might limit its parental care to relatively high quality offspring and eat the others. Or, an organism capable of repeated breeding in the same season might eat the survivors of a brood that was reduced by predators, and breed again. In this study, we manipulated brood size of convict cichlids by removing 0 % (control), 33 % (ER 33) or 66 % (ER 66) of the eggs spawned. The results show that smaller broods are more likely to be cannibalized by their parent(s) than are larger broods. Furthermore, pairs with reduced broods were preparing to respawn; female gonad weights were significantly higher in the brood-reduction groups than in the control group. In a second experiment, we show that the behaviour of the parents differed significantly between experimental groups; the ER 66 group performed less parental care than the control group. Moreover, females invested more parental effort than males. The results suggest that filial cannibalism may be a tactic used by parental convict cichlids to enhance their lifetime reproductive success.     

Cuckoldry incites cannibalism: male fish turn to cannibalism when perceived certainty of paternity decreases

Perceived certainty of paternity is expected to influence a male's behavior toward his offspring: if he is uncertain of his reproductive success with a current brood due to the presence of cuckolders, it may benefit him to invest instead in future reproduction. A decrease in perceived certainty of paternity incites filial cannibalism (the eating of one's own offspring) in some teleost fishes that provide parental care; however, no work has demonstrated that cannibalism increases proportionately with increased levels of cuckoldry. Here we show for the first time in a fish with no parental care that as the number of cuckolders at a spawning event increases, so does the probability that a male will cannibalize eggs. In field observations of Telmatherina sarasinorum, a small fish endemic to Sulawesi, Indonesia, males increased filial cannibalism behavior threefold in the presence of one cuckolder and nearly sixfold in the presence of two or more cuckolders. This suggests that males may use detection of cuckolders as an indication that the paternity of current offspring has been compromised.     

The effect of brood cannibalism on the operational sex ratio in parental teleost fishes

The operational sex ratio (OSR), an important determinant of the intensity of sexual selection, can vary with the population sex ratio and the potential reproductive rates of males and females. In parental teleosts, different forms of brood cannibalism, filial and non-kin, could affect the potential reproductive rates of males and females, and thereby the OSR. The direction of the effect of brood cannibalism on the OSR would depend on the sex of the cannibal and the parent. No evidence has been found that brood cannibalism can affect the OSR by altering the sex ratio.     

Experimental evolution of a more restrained clutch size when filial cannibalism is prevented in burying beetles Nicrophorus vespilloides

The overproduction of offspring is commonly associated with high hatching failure and a mechanism for dispensing with surplus young. We used experimental evolution of burying beetle populations Nicrophorus vespilloides to determine causality in these correlations. We asked does eliminating the mechanism for killing “spare” offspring cause the evolution of a more restrained clutch size and consequently select for reduced hatching failure? N. vespilloides typically overproduces eggs but kills 1^st instar larvae through partial filial cannibalism during brood care. We established replicate evolving populations that either could practice filial cannibalism (Full Care) or could not, by removing parents before their young hatched (No Care). After 20+ generations of experimental evolution, we measured clutch size and hatching success. We found that No Care females produced fewer eggs than Full Care females when allowed to breed on a small corpse, a finding not explained by differences in female quality. On larger corpses, females from both populations laid similar numbers of eggs. Furthermore, hatching success was greater in the No Care populations on small corpses. Our results suggest that the adaptive overproduction of offspring depends on a mechanism for eliminating surplus young and that killing offspring, in turn, relaxes selection against hatching failure.     

Avoidance of Filial Cannibalism in the Amphipod Gammarus pulex

Adult animals that cannibalise juvenile conspecifics may gain energy but also risk filial cannibalism, that is, consumption of their own offspring. However, individuals vary in the magnitude of the costs and benefits of cannibalism depending on factors such as their current energy reserves or the probability that they have offspring in the vicinity. They may therefore also vary in the extent to which they participate in cannibalism. This study investigated whether the sex or brooding status of adult amphipods (Gammarus pulex) influenced whether they participated in cannibalism of juveniles. For females carrying embryos within their brood pouch, we also investigated two hypotheses to explain the presence or absence of cannibalistic behaviour by determining whether cannibalism was correlated with factors that might reflect energy demands (body length, brood size), or that might reflect a temporal change in cannibalistic behaviour (corresponding to stage of brood development). All reproductive classes of adults participated in some level of juvenile cannibalism, but females carrying offspring at an advanced stage of development (close to emergence from the brood pouch) consumed significantly fewer juveniles than other groups. Females thus appear to significantly reduce cannibalism of juveniles concurrent with the time when their own eggs are hatching within the brood pouch, prior to the release of their offspring. Because the experiment tested female responses to unfamiliar juveniles, this reflects a temporal change in behaviour rather than a response to phenotypic recognition cues, although additional direct recognition cannot be ruled out. Brooding females with large brood sizes or large body lengths, which might have disproportionately greater energetic demands, were not more likely to cannibalise juveniles. We also noted that juveniles that survived in trials where cannibalism occurred were significantly more likely to be found at the water surface, suggesting a possible adaptation to escape cannibalistic adults. Overall, our results provide evidence that amphipods use indirect temporal cues to avoid filial cannibalism.     

Effects of a low oxygen environment on parental effort and filial cannibalism in the male sand goby, Pomatoschistus minutus

In fish, brood cycling parental males sometimes eat some orall of their eggs, a behavior termed filial cannibalism. Wetested predictions of filial cannibalism models related to thecost of parental care in the male sand goby, Pomatoschistusminutus, by increasing the parental effort (fanning expenditure)through reduced levels of dissolved oxygen to 39% in an experimentalgroup, whereas a control group had fully saturated water. Malesshowed both full-clutch cannibalism and partial-clutch cannibalismin both treatments. Giving the males one to three females tospawn with, we found that small clutches were completely eatenmore often than were larger ones, whereas partial-clutch cannibalismwas not affected by clutch size. Although treatment did notaffect filial cannibalism, it did affect a male's energy statesuch that males in the low oxygen treatment lost more body fat,indicating a greater fanning effort. This shows that males inthe low oxygen treatment allocated more energy to the presentbrood, potentially at the expense of future reproductive success.Our study strongly suggests that filial cannibalism in malesand gobies represents a strategic life-history decision asan investment in future reproductive success, and is not triggeredby a proximate need for food necessary for the male's own survival.Furthermore, males in the low oxygen treatment built nests withlarger entrances, and were less likely to rebuild their nestsafter destruction. Presumably, this makes fanning easier butthe nest more vulnerable to predators, suggesting a trade-offbetween fanning and nest defense.     

The evolution of filial cannibalism and female mate choice strategies as resolutions to sexual conflict in fishes

Filial cannibalism (the consumption of one's own viable offspring) is common among fish with paternal care. In this study, I use a computer simulation to study simultaneous evolution of male filial cannibalism and female mate choice. Under certain conditions, selection on parental males favors filial cannibalism. When filial cannibalism increases a male's probability to raise the current brood successfully, filial cannibalism also benefits the female. However, when egg eating is a male investment into future reproduction, a conflict between female and male interests emerges. Here I investigate how female discrimination against filial cannibals affects evolution of filial cannibalism and how different female choice criteria perform against filial cannibalism. The introduction of discriminating females makes the fixation of filial cannibalism less likely. I introduced three different female choice criteria: (1) females who could discern a male's genotype, that is, whether the male was going to eat eggs as an investment in future reproductive events; (2) energy-choosing females that preferred to mate with males who had enough energy reserves to live through the current brood cycle without consuming eggs; and (3) females that preferred to mate with already mated males, that is, males with eggs in their nest. Genotype choice never coexisted with filial cannibals at fixation and filial cannibals were unable to invade a population with genotype-choosing females. Energy choice was successful only when males had high energy reserves and were less dependent on filial cannibalism as an alternative energy source. The egg choosers frequently coexisted with the cannibals at fixation. When the female strategies were entered simultaneously, the most frequent outcome for low mate sampling costs was that both the cannibals and the egg choice was fixed and all other strategies went extinct. These results suggest that sexual conflicts may not always evolve toward a resolution of the conflict, but sometimes the stable state retains the conflict. In the present case, this was because the egg-preference strategy had a higher fitness than the other female strategies. The outcome of this simulation is similar to empirical findings. In fish with paternal care, male filial cannibalism and female preference for mates with eggs commonly co-occur.     

You eat what you are: personality‐dependent filial cannibalism in a fish with paternal care

Many animal parents invest heavily to ensure offspring survival, yet some eventually consume some or all of their very own young. This so‐called filial cannibalism is known from a wide range of taxa, but its adaptive benefit remains largely unclear. The extent to which parents cannibalize their broods varies substantially not only between species, but also between individuals, indicating that intrinsic behavioral differences, or animal personalities, might constitute a relevant proximate trigger for filial cannibalism. Using a marine fish with extensive paternal care, the common goby (Pomatoschistus microps), we investigated the influence of animal personality on filial cannibalism by assessing (1) behavioral consistency across a breeding and a nonbreeding context; (2) correlations between different breeding (egg fanning; filial cannibalism) and nonbreeding (activity) behaviors, and, in a separate experiment; (3) whether previously established personality scores affect filial cannibalism levels. We found consistent individual differences in activity across contexts. Partial filial cannibalism was independent of egg fanning but correlated strongly with activity, where active males cannibalized more eggs than less active males. This pattern was strong initially but vanished as the breeding season progressed. The incidence of whole clutch filial cannibalism increased with activity and clutch size. Our findings indicate that filial cannibalism cannot generally be adjusted independently of male personality and is thus phenotypically less plastic than typically assumed. The present work stresses the multidimensional interaction between animal personality, individual plasticity and the environment in shaping filial cannibalism.     

Parental fish change their cannibalistic behaviour in response to the cost-to-benefit ratio of parental care

Partial filial cannibalism, the act of cannibalizing some offspring, has been explained as a response to the high energetic cost of care. I tested this hypothesis by manipulating the cost-to-benefit ratio of care in the scissortail sergeant, Abudefduf sexfasciatus, a tropical damselfish with male care. Background egg mortality was lower than the incidence of cannibalism, confirming that males did not just dispose of dead eggs. Investment in the current brood affected future investment, because males forced to skip a brood cycle put more effort into courtship during the following cycle and obtained larger broods than did unmanipulated males. Any factor influencing the cost-to-benefit ratio of parental care should also affect the incidence of cannibalism. I reduced the cost of care by supplementary feeding and reduced the benefit of care by simulating a decrease in paternity certainty through simulated intrusions by non-nesting males. Supplementary feeding significantly reduced partial filial cannibalism by parental males, a result compatible with the hypothesis that eggs are consumed to cover the energetic costs of parental care. Cannibalism decreased regardless of whether males were fed with conspecific eggs or crabmeat. Cannibalism was only reduced but not fully eliminated by supplementary feeding, and residual levels of cannibalism after feeding were similar to the background rate of egg mortality. Simulated intrusions by non-nesting males led to an increase in filial cannibalism and a decrease in parental effort.     

Does time of the season influence filial cannibalism in the sand goby, <Emphasis Type="Italic">Pomatoschistus minutus</Emphasis>?

According to life-history theory, filial cannibalism by fish that breed over one season only should be more beneficial early than late in the season if they eat eggs to invest energy into later clutches. Also, filial cannibalism may be more costly late in the season if finding ripe females for replacing eaten eggs is harder then. On the other hand, offspring hatching early may have a competitive advantage over fry hatching late and hence provide higher fitness to the parent. Using data collected over three successive years, I tested if sand goby males are more prone to eat of their eggs early than late in the reproductive season. I found no difference in the amount of eggs eaten or in the frequency of males eating the whole clutch between early and late in the season. Furthermore, there was no difference in the frequency of males who ate parts of their clutches, early compared to late. This might reflect a trade-off between quality (early hatching offspring) and quantity (producing as many offspring as possible over a long reproductive season). If so, the lack of seasonal pattern of filial cannibalism found in sand gobies might be the result of opposing selection pressures.     

Filial cannibalism improves survival and development of beaugregory damselfish embryos

Cannibalism of small numbers of offspring by a parent has been proposed as an adaptive parental strategy, by providing energy to support parental care. However, there are few empirical studies to support this hypothesis. We conducted field and laboratory experiments to investigate partial filial cannibalism in Stegastes leucostictus, a coral reef fish with paternal care. Partial cannibalism was shown to be common, and males were found to remove developing embryos from throughout a clutch in a random pattern, rather than in the more aggregated pattern seen during embryo predation. Males that received a diet supplement grew faster than control males, but did not engage in less cannibalism. Also, males did not concentrate cannibalism on early embryonic stages with the highest energetic value. Experimental reduction of embryo densities was found to significantly increase embryo development rate and survival from egg deposition to hatching, and experimental reduction of oxygen levels significantly increased rates of partial filial cannibalism by males. Artificial spawning sites with low oxygen levels were avoided by spawning females, and cannibalism rates by males were higher. We propose that partial filial cannibalism serves as an adaptive parental strategy to low oxygen levels in S. leucostictus by increasing the hatching success of embryos.     

Paternity assurance before and after fertilization by male burying beetles (Nicrophorus quadripunctatus)

Parental care requires a large investment of time and energy. This can reduce future parental survival and opportunities for mating. Because males are usually more uncertain of their parentage with respect to the caring of offspring than are females, the reduction in reproductive success is thought to be greater in males. Therefore, males are under selection to ensure paternity of the offspring for which they care. Males can increase paternity before and after fertilization. Before fertilization, males can increase paternity by increasing their competitive ability for fertilization. After fertilization, males can increase paternity by cannibalizing unrelated offspring. Here, we investigated the stage at which male burying beetles, Nicrophorus quadripunctatus, increase their paternity by evaluating the number of offspring sired by a nursing male in asynchronously hatched broods in relation to hatching time. We found that nursing males assure a very high level of the paternity of hatching offspring. We also found that the paternity of non-nursing and nursing males remained constant across hatching time within a brood, indicating that it is unlikely that filial cannibalism plays a role in increasing the paternity of offspring. We concluded that ensuring paternity before fertilization is more important in increasing the paternity of offspring.     

Water mold infection but not paternity induces selective filial cannibalism in a goby

Many animals heavily invest in parental care but still reject at least some of their offspring. Although seemingly paradoxical, selection can favor parents to neglect offspring of particularly low reproductive value, for example, because of small survival chances. We here assess whether filial cannibalism (FC), where parents routinely eat some of their own young, is selective in response to individual offspring reproductive value. We performed two independent laboratory experiments in the common goby (Pomatoschistus microps) to test whether caring fathers preferentially cannibalize eggs of a given infection history and paternity. While males did not discriminate kin from nonkin eggs, they consumed significantly more eggs previously exposed to water mold compared to uninfected eggs. Our findings clearly show that parents differentiate between eggs based on differences in egg condition, and thus complement the prevailing view that FC arises for energetic reasons. By preventing the spread of microbial infections, the removal of molded eggs can constitute an important component of parental care and may represent a key driver of selective FC in a wide array of parental fish.     

Big and tall parents have more sons: further generalizations of the Trivers-Willard hypothesis

This paper proposes the generalized Trivers-Willard hypothesis (gTWH), which suggests that parents who possess any heritable trait which increases male reproductive success at a greater rate than female reproductive success in a given environment will have a higher-than-expected offspring sex ratio, and parents who possess any heritable trait which increases female reproductive success at a greater rate than male reproductive success in a given environment will have a lower-than-expected offspring sex ratio. Since body size (height and weight) is a highly heritable trait which increases male (but not female) reproductive success, the paper hypothesizes that bigger and taller parents have more sons. The analysis of both surviving children and recent pregnancies among respondents of the National Child Development Survey and the British Cohort Survey largely supports the hypothesis.     

Violent men have more sons: further evidence for the generalized Trivers-Willard hypothesis (gTWH)

The generalized Trivers-Willard hypothesis (gTWH) [Kanazawa, S., 2005a. Big and tall parents have more sons; further generalizations of the Trivers-Willard hypothesis. J. Theor. Biol. 235, 583-590] proposes that parents who possess any heritable trait which increases the male reproductive success at a greater rate than female reproductive success in a given environment have a higher-than-expected offspring sex ratio, and parents who possess any heritable trait which increases the female reproductive success at a greater rate than male reproductive success in a given environment have a lower-than-expected offspring sex ratio. One heritable trait which increases the reproductive success of sons significantly more than that of daughters in the ancestral environment is the tendency toward violence and aggression. I therefore predict that violent parents have a higher-than-expected offspring sex ratio (more sons). The analysis of both American samples and a British sample demonstrates that battered women, who are mated to violent men, have significantly more sons than daughters.     

Beautiful parents have more daughters: a further implication of the generalized Trivers-Willard hypothesis (gTWH)

The generalized Trivers-Willard hypothesis (gTWH) [Kanazawa, S., 2005. Big and tall parents have more sons: further generalizations of the Trivers-Willard hypothesis. J. Theor. Biol. 235, 583-590) proposes that parents who possess any heritable trait which increases the male reproductive success at a greater rate than female reproductive success in a given environment will have a higher-than-expected offspring sex ratio, and parents who possess any heritable trait which increases the female reproductive success at a greater rate than male reproductive success in a given environment will have a lower-than-expected offspring sex ratio. One heritable trait which increases the reproductive success of daughters much more than that of sons is physical attractiveness. I therefore predict that physically attractive parents have a lower-than-expected offspring sex ratio (more daughters). Further, if beautiful parents have more daughters and physical attractiveness is heritable, then, over evolutionary history, women should gradually become more attractive than men. The analysis of the National Longitudinal Study of Adolescent Health (Add Health) confirm both of these hypotheses. Very attractive individuals are 26% less likely to have a son, and women are significantly more physically attractive than men in the representative American sample.