Can evolutionary and ecological dynamics operating at one level of the biological hierarchy affect the dynamics and structure at other levels? In social insects, strong hostility towards unrelated individuals can evolve as a kin-selected counter-adaptation to intraspecific social parasitism. This aggression in turn might cause intraspecific competition to predominate over interspecific competition, permitting coexistence with other social insect species. In other words, kin selection—a form of intra-population dynamics—might enhance the species richness of the community, a higher-level structure. The converse effect, from higher to lower levels, might also operate, whereby strong interspecific competition may limit the evolution of selfish individual traits. If the latter effect were to prove more important, it would challenge the common view that intra-population dynamics (via individual or gene selection) is the main driver of evolution.
The 1960s was the Jazz Age of evolutionary ecology. In 1962, group selection was set out as an explicit mechanism for the evolution of altruism and dispersal behaviour of animals . Just four years later this idea was logically refuted by G. C. Williams . In this interval, W. D. Hamilton [3,4] proposed kin selection, which eventually became widely recognized as an alternative explanation for altruism.
The central theoretical ground for Williams's dismissal of group selection was the overwhelmingly faster turnover of individuals than of populations: group selection is too weak in the face of opposing individual selection. This and related discussions  directed evolutionary ecologists to focus on dynamics within a population, such as individual selection and sexual selection, to understand adaptation. At the same time, the main focus of behavioural ecologists shifted to individual-level phenomena such as social and mating behaviours. The effect of intra-population evolutionary dynamics on phenomena that can emerge at higher levels in the biological hierarchy, such as population regulation, community structure and ecosystem functions, has not been intensively studied until recently . However, if Williams's thesis—that intra-population evolutionary dynamics dominates inter-population dynamics in shaping individual organisms' traits—is generally correct, one could imagine that the evolutionary dynamics within populations has a strong influence on patterns at higher levels of the biological hierarchy.
Sexual selection can influence community structure . Strong inter-male competition can enhance male mating behaviour towards females of closely related species as a by-product . Such erroneous mating attempts have a fitness cost for those females [7,9]. Furthermore, this cost has a positive frequency dependency (the higher the ratio of heterospecific males to conspecific males, the lower the female fitness), and species that suffer such higher costs can quickly go extinct . This idea may be generalized as follows: strong intra-sexual selection can hinder the coexistence of closely related species and diminish species richness, regardless of their degree of similarity in resource use or niche . However, sexual selection might also have the opposite effect, because strong mate choice may enhance species richness through sympatric speciation . Thus, sexual selection can have disruptive effects on species richness depending on whether it is intra- or intersexual.
The evolution of altruism is another major topic in behavioural ecology. In this context, too, kin selection, a form of intra-population dynamics, might influence community structure. Let us begin with the issue of invasive exotic ants that has attracted the attention of both applied and basic ecologists. One interesting characteristic shared by many invasive ants is unicoloniality, in which ants lose colony-membership discrimination, and the population as a whole becomes a huge supercolony, which can even become discontinuously spread over populations. In fact, a single supercolony of the Argentine ant (Linepithema humile) inhabits Europe over a distance of 6000 km from Portugal to Italy and furthermore extends overseas to California, Australia and Hawaii [12,13]. As a result of the freedom of individuals to move between nests within the supercolony, nest-mate relatedness, defined as the genetic similarity of nest-mates in comparison to the background population average, has decreased to zero , and yet their cooperative society is maintained, in what is regarded as a major challenge to kin selection theory .
The bottleneck hypothesis  explains this phenomenon as follows. Exotic organisms often experience a genetic bottleneck on their introduction to new range, which reduces their intra-population genetic diversity. Ants are considered to discriminate the colony membership of individuals by using genetically based chemical labels, such as cuticular hydrocarbons . Ant populations after a bottleneck exhibit greatly reduced genetic variation in the labels. This makes colony discrimination difficult, and as a result colonies fuse. One community-level consequence of a bottleneck is the invasiveness of alien ants. Ants are generally believed to be aggressive animals whose population is structured into many mutually hostile colonies by fierce aggression, especially towards conspecific aliens . Under unicoloniality, the cost of intraspecific competition among colonies is reduced, allowing increases in population density. Furthermore, interspecific competition predominates over intraspecific competition, leading to the exclusion of native ants by exotic ants, as community ecology theory [18,19] predicts.
Arguments on the validity of the bottleneck hypothesis  are outside the scope of this paper. Instead, let us focus on an implication of the hypothesis. Note that the bottleneck hypothesis is a proximate explanation in which unicoloniality is regarded as a maladaptive trait caused by the human-mediated event (bottleneck). This hypothesis does not focus on the question of how the native ant communities, with many ant species usually coexisting, have evolved. It might, however, implicitly assume that the coexistence of native ant species before the invasion of a unicolonial alien is maintained by strong intraspecific competition that predominates over interspecific competition. The hostility among conspecific ant colonies is an established fact, albeit with some exceptions . However, the adaptive significance of this hostility, which might determine the community structure, is not well understood yet. Hostility towards aliens must be a defence against resource exploitation. Among ants, resources can be divided into ecological resources, such as nest sites, food and territory, and social resources, such as the cooperative workforce. Although some ants colonies do defend territories clearly to maintain exclusive foraging rights, absolute occupation by ants of a territory—an area always occupied more or less exclusively by a colony, with the exclusivity maintained by means of aggression—is rare. In many species, aggression is restricted in both space and time, and possibly the most generally observable aggression occurs when a worker meets a conspecific alien near the nest entrance. By contrast, there is no example of an ant that is non-aggressive towards a conspecific alien entering the nest yet aggressive towards an alien encountered far from the nest. This suggests that the defence of social resources is more important for ants than the defence of ecological resources, as the former is generally threatened by exploitation by conspecifics. Colony recognition and aggression towards distantly related aliens could form a counter-strategy against such a threat . Indeed, there are many examples of conspecific social parasites that intrude into an unrelated colony, breaking through the host's defences and selfishly reproducing by exploiting the social resources of the host colony [22–24]. In the most striking case, some parthenogenetic worker lineages of the Cape honeybee (Apis mellifera capensis) became social-parasitic and migrated among neighbouring captive colonies of the subspecies Apis mellifera scutellata, causing mass extinction of the captive population in the host colonies .
The most important concern for helper workers of social insects is the protection of their indirect (inclusive) fitness return through their cooperative efforts against parasitic exploitation. This implies that the primary adaptive significance of the repulsion between conspecific colonies may be the maintenance of intra-colonial relatedness; that is, aggression is a kin-selected trait . This strong aggression between different families (colonies) might give rise to the predominance of intraspecific competition (or population regulation) over interspecific competition as a by-product, allowing the coexistence of different ant species. One important contention of this hypothesis is that the ultimate mechanism for species coexistence among ants might not be differences in resource niches, as is often believed [17,25], but rather kin selection, a force operating within each member species of the community. Future studies should focus on the relative importance of this hypothesis and others, such as niche partitioning and neutrality , for ant species coexistence both empirically and theoretically. There is one crucial question to be solved concerning the kin-selected coexistence hypothesis: how can genetic variation in the colony recognition label evolve, overcoming the predicted positive frequency-dependent selection by which a colony with a rare mutant label allele will be more frequently attacked by conspecifics and thus have low fitness (Crozier's paradox)? [27,28].
The scope of kin selection has already been extended to phenomena lower than the individual level of the biological hierarchy, such as differential expression of paternal and maternal genes [29,30]. I hope that this essay will shed light on another direction of scope extension, enabling a new interaction between behavioural ecologists and community ecologists. Hamilton contended that landscape-level phenomena such as clouds  and autumn leaves  can result from adaptive strategies of individual organisms. It should not be too radical to consider that kin selection, his best-known theory, can be extended to explain macro-level phenomena such as community structure.
I conclude that the interaction of selective dynamics across the biological hierarchy will be an important subject in evolutionary ecology. Interestingly, some authors have started theoretically to discuss the effect of higher-level selective dynamics on lower ones [33,34], contrary to my view on the direction of the effect. Adaptive dynamics within a population, particularly those related to social interaction, can lead to the evolution of selfish traits that reduce average population fitness, i.e. ‘tragedy of commons’. Such a population or species is more likely to extinct  and/or lose interspecific competition . Thus, species-selection limits the evolution of extremely selfish traits. I consider this as another possible consequence of the interaction across the biological hierarchy. Given the possible presence of interaction across the biological hierarchy, the next question would be which effect, ‘lower to higher’ or ‘higher to lower’, is generally more important. If empirical evidence were to support the relative importance of the higher to lower effect, it implies that the power of intra-population dynamics is not as strong as of dynamics at higher levels in shaping both community structure and individual traits. It might also imply that differential extinction can have a strong impact on the evolution of individual characteristics, including altruism [1,33,34]. If so, evolutionary ecologists will have to come back to the consensus in the 1960s before Williams's dismissal of group selection and start another 50-year round of discussions.
This study was supported in part by KAKENHI (23405011 and 25660266).
I thank S. Dobata for discussion.
One contribution of 12 to the Special Feature ‘50 Years on: the legacy of William Donald Hamilton’ organized by Joan Herbers and Neil Tsutsui.
- Received May 28, 2013.
- Accepted July 5, 2013.
- © 2013 The Author(s) Published by the Royal Society. All rights reserved.