The Evolution of Cooperation
Cooperation is an evolutionary puzzle. While a simple reading of natural selection as ‘survival of the fittest’ predicts that individuals will benefit from selfish, competitive behaviour, cooperation is not only commonplace within nature, but is of profound evolutionary and social significance. Central to understanding the evolution of cooperation are the situations under which conflicts of interest arise between members of societies, and how these conflicts are resolved. My current research investigates the most fundamental of these conflicts; conflict over reproduction. If cooperation acts to increase the reproductive productivity of a group, how are the profits of cooperation to be shared among group members? Under what circumstances can reproduction become monopolised by a subset of individuals? And what are the impacts of differences in reproductive success on cooperation? Whilst these questions have an obvious human interest, they apply in all societies, from social insects through to fish, birds and mammals.
Our work has often used banded mongooses Mungos mungo to answer these questions. The banded mongoose is an ideal species in which to investigate the impact of reproductive conflict on the evolution of cooperative societies because the distribution of reproduction between group-members (or level of ‘reproductive skew’) varies widely over time. My work has addressed five key questions:
1. What explains variation in reproductive skew among females? While reproductive skew has been a focus of research for over twenty years, it has not yet been possible to explain the distribution of subordinate breeding either between or within species. Previous attempts have often assumed that subordinate females breed when dominant females grant them reproductive concessions to ensure that they help to rear the dominants’ offspring, or they breed when dominants are incapable of preventing the subordinates from conceiving or rearing young. Our work (in colloboration with Professor Mike Cant, Dr Sarah Hodge, Dr Emma Vitikainen, Dr Harry Marshall, Dr Jenni Sanderson, Dr Faye Thompson and Dr Matt Bell) supports an alternative view that has received relatively little attention: reproductive skew appears to be determined by a trade-off between the costs and benefits to dominants of suppressing subordinate reproduction. In the banded mongoose, dominant females suppress subordinate reproduction by aggressively evicting subordinates from the group en masse, who then abort their litters (https://doi.org/10.1098/rspb.2015.2607). Dominants suffer a substantial cost of evicting: the disruption and stress caused by an eviction can result in reduced weight, growth and survival of the evictor’s pups (https://doi.org/10.1098/rspb.2011.1093). Given that evictions are costly, they should only occur when the benefits of doing so are large. We have shown that evictions are ecologically contingent, and occur when food is scarce (https://doi.org/10.1093/beheco/ars008) and group sizes are large (https://doi.org/10.1073/pnas.1312626111). At these times, the benefit of evicting subordinates (that may produce competing offspring) is much greater than when food is plentiful. Interestingly, dominant females target close kin for eviction, which may relate to the costs and benefits to subordinates of resisting eviction (https://doi.org/10.1073/pnas.1612235114). Reproductive competition is not limited to eviction, and stress levels experienced during pregnancy may also regulate subordinate reproduction, suggesting that dominants may be able to suppress subordinates without evicting them (https://doi.org/10.1038/srep20013). Females also appear to moderate the pre-natal growth of their offspring according to the likely level of competition they will experience after birth (https://doi.org/10.1038/srep20013). Together, these results suggest that much of the variation in reproductive skew both within and between social species is determined by adaptive adjustment in the effort invested in suppression by dominants, particularly in response to variation in resource availability.
2. What explains variation in reproductive skew among males? Reproductive skew among males has traditionally been considered a product of intrasexual competition and escalating benefits of dominance. Our research suggests this view is incomplete, and highlights the additional importance of interactions between the sexes in determining skew amongst males (https://doi.org/10.1016/j.anbehav.2010.06.025).
3. What are the consequences of variation in reproductive skew on cooperation? Our work has demonstrated that variation in reproductive skew is likely to generate differences in cooperative behaviour. Pregnant females give birth (usually on the same night) in an underground den and the resulting litter is raised communally with the vast majority of group-members contributing to pup care. We found that females that were mothers of members of the communal litter put more effort into raising pups than females that were not, especially when food was scarce and the potential costs of caring for pups were greater (https://doi.org/10.1016/j.anbehav.2012.03.005). However, mothers (or fathers) do not appear capable of recognising, and hence caring for, their own pups (https://doi.org/10.1098/rspb.2016.2384). This lack of offspring-recognition is unusual for a mammal and is likely to arise because cues to maternity are scrambled by the extreme birth synchrony.
4. How are traits inherited in cooperative societies? Behavioural traits can be inherited genetically, or via social learning, but these inheritance routes are often difficult to distinguish as parents are usually the source of both genetic and cultural information. Banded mongooses provide an excellent opportunity to detangle these inheritance routes as pups are primarily cared for in 1:1 relationships by 'escorts' who act as social role-models but are rarely the parents of the offspring they care for. We found that pups inherit lifelong foraging niches from their escorts (https://doi.org/10.1016/j.cub.2018.05.001). In contrast, cooperative personalities are not inherited from the escorts but do show genetic inheritance. Understanding the situations under which traits are socially versus genetically inherited will reveal much about the role of culture in the evolution of animal societies.
4. How does cooperation evolve? Whilst there has been much focus on early-life helping behaviour (i.e. cooperative breeding), we have a relatively poor understanding of helping behaviour later in live, for example post-reproductive grandmothers helping their adult offspring and grand-offspring. We have worked with phylogeneticist Dr Kevin Arbuckle to investigate patterns of post-reproductive lifespan (termed menopause in humans) across mammals (https://doi.org/10.1098/rsbl.2015.0992,). As well as occurring in humans, long post-reproductive lifespans (PRLS) are found in a handful of toothed whales, including the orca and short-finned pilot whale. We found that species-wide PRLS doesn't occur in long-finned pilot whales, despite them showing the similar relatedness patterns to whales that do have PRLS. However, female long-finned pilot whales with a lot of philopatric offspring are less likely to reproduce in later life, suggesting that a small number of older females may switch from reproducing to helping their existing offspring in later life (https://doi.org/10.1093/beheco/arz211). Work in progress by MRes student Jack McCormack suggests that female long-finned pilot whales may continue to reproduce in later life due to a low level of competition between their offspring and the offspring of their philopatric daughters. A large-scale study, combining modelling and empirical approaches, is currently underway alongside researchers at the University of Exeter, investigating how the relatedness structures of mammalian societies change as individuals age, and how this in turn impacts on the evolution of helping and harming behaviours.
Our work has often used banded mongooses Mungos mungo to answer these questions. The banded mongoose is an ideal species in which to investigate the impact of reproductive conflict on the evolution of cooperative societies because the distribution of reproduction between group-members (or level of ‘reproductive skew’) varies widely over time. My work has addressed five key questions:
1. What explains variation in reproductive skew among females? While reproductive skew has been a focus of research for over twenty years, it has not yet been possible to explain the distribution of subordinate breeding either between or within species. Previous attempts have often assumed that subordinate females breed when dominant females grant them reproductive concessions to ensure that they help to rear the dominants’ offspring, or they breed when dominants are incapable of preventing the subordinates from conceiving or rearing young. Our work (in colloboration with Professor Mike Cant, Dr Sarah Hodge, Dr Emma Vitikainen, Dr Harry Marshall, Dr Jenni Sanderson, Dr Faye Thompson and Dr Matt Bell) supports an alternative view that has received relatively little attention: reproductive skew appears to be determined by a trade-off between the costs and benefits to dominants of suppressing subordinate reproduction. In the banded mongoose, dominant females suppress subordinate reproduction by aggressively evicting subordinates from the group en masse, who then abort their litters (https://doi.org/10.1098/rspb.2015.2607). Dominants suffer a substantial cost of evicting: the disruption and stress caused by an eviction can result in reduced weight, growth and survival of the evictor’s pups (https://doi.org/10.1098/rspb.2011.1093). Given that evictions are costly, they should only occur when the benefits of doing so are large. We have shown that evictions are ecologically contingent, and occur when food is scarce (https://doi.org/10.1093/beheco/ars008) and group sizes are large (https://doi.org/10.1073/pnas.1312626111). At these times, the benefit of evicting subordinates (that may produce competing offspring) is much greater than when food is plentiful. Interestingly, dominant females target close kin for eviction, which may relate to the costs and benefits to subordinates of resisting eviction (https://doi.org/10.1073/pnas.1612235114). Reproductive competition is not limited to eviction, and stress levels experienced during pregnancy may also regulate subordinate reproduction, suggesting that dominants may be able to suppress subordinates without evicting them (https://doi.org/10.1038/srep20013). Females also appear to moderate the pre-natal growth of their offspring according to the likely level of competition they will experience after birth (https://doi.org/10.1038/srep20013). Together, these results suggest that much of the variation in reproductive skew both within and between social species is determined by adaptive adjustment in the effort invested in suppression by dominants, particularly in response to variation in resource availability.
2. What explains variation in reproductive skew among males? Reproductive skew among males has traditionally been considered a product of intrasexual competition and escalating benefits of dominance. Our research suggests this view is incomplete, and highlights the additional importance of interactions between the sexes in determining skew amongst males (https://doi.org/10.1016/j.anbehav.2010.06.025).
3. What are the consequences of variation in reproductive skew on cooperation? Our work has demonstrated that variation in reproductive skew is likely to generate differences in cooperative behaviour. Pregnant females give birth (usually on the same night) in an underground den and the resulting litter is raised communally with the vast majority of group-members contributing to pup care. We found that females that were mothers of members of the communal litter put more effort into raising pups than females that were not, especially when food was scarce and the potential costs of caring for pups were greater (https://doi.org/10.1016/j.anbehav.2012.03.005). However, mothers (or fathers) do not appear capable of recognising, and hence caring for, their own pups (https://doi.org/10.1098/rspb.2016.2384). This lack of offspring-recognition is unusual for a mammal and is likely to arise because cues to maternity are scrambled by the extreme birth synchrony.
4. How are traits inherited in cooperative societies? Behavioural traits can be inherited genetically, or via social learning, but these inheritance routes are often difficult to distinguish as parents are usually the source of both genetic and cultural information. Banded mongooses provide an excellent opportunity to detangle these inheritance routes as pups are primarily cared for in 1:1 relationships by 'escorts' who act as social role-models but are rarely the parents of the offspring they care for. We found that pups inherit lifelong foraging niches from their escorts (https://doi.org/10.1016/j.cub.2018.05.001). In contrast, cooperative personalities are not inherited from the escorts but do show genetic inheritance. Understanding the situations under which traits are socially versus genetically inherited will reveal much about the role of culture in the evolution of animal societies.
4. How does cooperation evolve? Whilst there has been much focus on early-life helping behaviour (i.e. cooperative breeding), we have a relatively poor understanding of helping behaviour later in live, for example post-reproductive grandmothers helping their adult offspring and grand-offspring. We have worked with phylogeneticist Dr Kevin Arbuckle to investigate patterns of post-reproductive lifespan (termed menopause in humans) across mammals (https://doi.org/10.1098/rsbl.2015.0992,). As well as occurring in humans, long post-reproductive lifespans (PRLS) are found in a handful of toothed whales, including the orca and short-finned pilot whale. We found that species-wide PRLS doesn't occur in long-finned pilot whales, despite them showing the similar relatedness patterns to whales that do have PRLS. However, female long-finned pilot whales with a lot of philopatric offspring are less likely to reproduce in later life, suggesting that a small number of older females may switch from reproducing to helping their existing offspring in later life (https://doi.org/10.1093/beheco/arz211). Work in progress by MRes student Jack McCormack suggests that female long-finned pilot whales may continue to reproduce in later life due to a low level of competition between their offspring and the offspring of their philopatric daughters. A large-scale study, combining modelling and empirical approaches, is currently underway alongside researchers at the University of Exeter, investigating how the relatedness structures of mammalian societies change as individuals age, and how this in turn impacts on the evolution of helping and harming behaviours.