Search

Johnstone and Cant introduce menopause and the phenomenon of an extended post-reproductive life found in humans and very few other mammals.

Summary Why females of some species cease ovulation prior to the end of their natural lifespan is a long-standing evolutionary puzzle [1–4]. The fitness benefits of post-reproductive helping could in principle select for menopause [1, 2, 5], but the magnitude of these benefits appears insufficient to explain the timing of menopause [6–8]. Recent theory suggests that the cost of inter-generational reproductive conflict between younger and older females of the same social unit is a critical missing term in classical inclusive fitness calculations (the "reproductive conflict hypothesis" [6, 9]). Using a unique long-term dataset on wild resident killer whales, where females can live decades after their final parturition, we provide the first test of this hypothesis in a non-human animal. First, we confirm previous theoretical predictions that local relatedness increases with female age up to the end of reproduction. Second, we construct a new evolutionary model and show that given these kinship dynamics, selection will favor younger females that invest more in competition, and thus have greater reproductive success, than older females (their mothers) when breeding at the same time. Third, we test this prediction using 43 years of individual-based demographic data in resident killer whales and show that when mothers and daughters co-breed, the mortality hazard of calves from older-generation females is 1.7 times that of calves from younger-generation females. Intergenerational conflict combined with the known benefits conveyed to kin by post-reproductive females can explain why killer whales have evolved the longest post-reproductive lifespan of all non-human animals.

SummaryThe study of animal signals began in earnest with the publication in 1872 of Charles Darwin’s The Expressions of the Emotions in Man and Animals, which laid the basis for a comparative study of signals across all animals, including humans. Yet even before Darwin, the exceptional diversity of animal signals has gripped the attention of natural historians and laymen alike, as these signals represent some of the most striking features of the natural world. Structures such as the long ornamented tail of the peacock, the roaring sounds of howler monkeys, audible kilometers away, and the pheromone trails laid by ants to guide their nestmates to resources are each examples of animal signals (Figure 1). Indeed, because signals evolved for the purpose of communicating (Box 1), their prominence can be hard for even a casual observer to overlook. Animal signals therefore raise many scientific questions: What are their functions? What information do they transmit? How are they produced? And why did they evolve?

The grouping behaviour of fish is a widespread phenomenon of high biological significance but little is known as to how consistent individual behavioural differences may affect group joining preferences. When given the option to join either a shy or a bold shoal of three-spined sticklebacks, Gasterosteus aculeatus, both shy and bold individuals showed a strong preference for associating with bold fish. Personality type interacted with individual hunger levels to affect the extent of association, suggesting important strategy variation by focal fish in a competitive foraging environment. Furthermore, shoals modified their behaviour in relation to the focal individual. Individual behavioural differences were shown to have a complex role in influencing association preferences as well as driving previously unrecognized behavioural modifications in foraging groups.

In communally nesting species such as mice, nursing females may be confronted with alien and own young in mixed litters, and offspring may be fostered by either own or alien mothers. Because of the costly nature of maternal investment and offspring solicitation behaviour, maternal provisioning is expected to favour own young. While previous studies have investigated kin recognition between mother and pups with inconclusive results, it remains unexplored whether any such ability yields a fitness advantage for own pups when nursed in a mixed litter. Moreover, nursing females may have evolved strategies to effect a reduction in fitness costs associated with nursing alien pups, for example, reducing their maternal effort. In this study, we show that young mice procure an advantage by gaining relatively more weight than alien pups when nursed together in a mixed litter by their own mothers. However, we did not detect an effect of offspring or maternal genotype on this difference in weight gain. Furthermore, we show that mothers provide overall fewer resources to mixed litters than to unmixed litters and that two mouse strains differ in the magnitude of this reduction. We suggest that a combination of sibling competition and maternal strategies to reduce costs of nursing unrelated young may be the underlying causes of these findings.

Many communal breeders are characterized by a conflict over who gets to reproduce, with dominant individuals often claiming the largest share of reproduction in the group. How do dominants control breeding in these species? Although infanticide has often been invoked as a means of control, previous theoretical work on indiscriminate killing of young did not support this idea. There is, however, increasing evidence from field studies in both vertebrates and insects that infanticidal individuals can discriminate between their own offspring and those of other group members, and thus avoid the risk of accidentally killing their own progeny. In a simple game-theoretical model we demonstrate that the capacity for discriminate infanticide can promote high reproductive skew even though few or no offspring are actually killed. When discrimination is good and offspring are cheap to produce, the threat of infanticide prevents the subordinate cobreeder from adding many young to the joint brood, and no killing need occur. High levels of infanticide tend to occur only when discrimination is poor, costs of offspring production are low and/or relatedness is low.

Recent signalling models have shown that honest, cost-free communication between relatives can be stable. Moreover, cost-free signalling equilibria are in some cases more efficient than costly equilibria. However, we show that they are also relatively uninformative, particularly when relatedness between signaller and receiver is low. We explore the trade-off between signal cost and information, and further demonstrate that incorporating competition among signallers into a model of communication between relatives can reduce the propensity of any one signaller to display. As a result, there is a general increase in the amount of broadcast information in a non-costly signal with increasing competitor number.

In communally breeding animals, there is an evolutionary conflict over the partitioning of reproduction within the group. If dominant group members do not have complete control over subordinate reproduction, this conflict may favour the evolution of infanticidal behaviour (by either subordinates or dominants or both). Elimination of offspring, however, is likely to be constrained by the difficulty of discriminating between an individual's own progeny and those of cobreeders. Here, we develop an evolutionarily stable strategy (ESS) model of reproductive partitioning, which demonstrates that killing of young can be favoured, even if such discrimination is not possible. The model predicts that infanticide will typically be associated with elevated levels of offspring production, and is most likely to prove evolutionarily stable when the coefficient of relatedness between cobreeders is low, and offspring are cheap to produce. The effect of infanticide is to release subordinates from the reproductive restraint they would otherwise be forced to exercise, leading to reduced reproductive skew. When infanticide is possible, addition of numerous young to the joint brood will not lower overall productivity, because progeny in excess of the most productive brood size are eliminated. Subordinates are thus free to contribute more young to the brood than would otherwise be the case. In addition, we show that the possibility of infanticide may influence the pattern of reproduction within a group even if no offspring are actually killed at equilibrium. Copyright 1999 The Association for the Study of Animal Behaviour.

Throughout the animal kingdom, distinctive behaviour by offspring commonly precedes and accompanies their provisioning by parents. Here, we assess empirical support for the recent theory that begging advertises offspring need, that parents provision young in relation to begging intensity, and that the apparently costly nature of begging ensures the reliability of the signal. While there is some support for the predictions of honest signalling models, empirical work has also revealed a host of complexities (such as the use of multiple signals) that existing theoretical analyses have only begun to address.

Animal displays are often complex, involving a variety of different visual, auditory and/or olfactory components. This observation poses a problem for models of signalling based on the handicap principle, which predict that displays generally serve to advertise quality, because it is not obvious why honest advertisement should require multiple signals. One possible explanation is that complex, multi-component displays provide information about many different aspects of the quality or condition of the signaller. Here, a game-theoretical model of signalling is described in which multiple signals serve to advertise multiple qualities in this way. When many different kinds of signal are available, there can be no guarantee that a particular signal will be less costly for a signaller of higher overall value. Nevertheless, the model demonstrates that honest signalling using multiple displays can be stable; multiple signal equilibria exist at which receivers acquire accurate information about the overall value of signallers. It is also shown that, at such equilibria, there need be no one-to-one relationship between signals and qualities. Even if the cost of a particular signal trait depends only on one particular quality, its expression is likely to be influenced by other qualities as well.

The last 20 years have seen the accumulation of a large body of information on selfish genetic elements — genes that act to further their own evolutionary interests at a cost to the individual (genome) bearing them. During the last few years, a growing number of authors have suggested that the intragenomic conflict these elements create is not just an intriguing example of natural selection in action, but a driving force behind the evolution of genetic systems. A host of phenomena, from exquisite details of gene expression to the evolution of crossing over, from the existence of syncytia during gametogenesis to the amount of DNA present in eukaryotes and the existence of multicopy genes, may all be explicable as the result of conflict within the nuclear genome.

An analysis of Maynard Smith's two-player, ESS model of biological signalling, the "Sir Philip Sidney game", is presented. The stable strategies of the players in this game are shown to satisfy the conditions of Zahavi's handicap principle. At equilibrium, signals are honest, costly, and costly in a way that is related to the true quality revealed. Further analysis reveals that the level of cost required to maintain stability is inversely related to the degree of relatedness between the players. It therefore seems likely that stable biological signalling systems will feature lower signalling costs when communication occurs between relatives. A three-player, extended version of the model is investigated, in which signals are passed via an intermediate, or "messenger". It is shown that this destabilizes the signalling system, and leads to increased signalling costs. This result suggests that "kin conflict" theories of the evolution of the endosperm in flowering plants require further refinement. The introduction of a novel resource acquisition tissue, which mediates parent-offspring interaction during development, cannot be assumed to limit parent-offspring conflict simply because it carries an extra copy of the maternally inherited genes. The ability to add such complications to the Sir Philip Sidney game and still obtain solutions makes it a very useful modelling tool.