Inclusive fitness of ‘kissing cousins’: new evidence of a role for kin selection in the evolution of extra-pair mating in birds


Kevin Oh, Fax: 607 254 4320; E-mail:


Social monogamy is nearly ubiquitous across avian taxa, but evidence from a proliferation of studies utilizing molecular paternity analysis suggests that sexual monogamy is the rare exception rather than the rule (Griffith et al. 2002). Efforts to explain the prevalence of extra-pair paternity (EPP) have largely focused on the potential fitness benefits for offspring genetic quality, as females are less likely to benefit directly from seeking extra-pair mates. In particular, there has been considerable interest in the degree to which EPP may represent an adaptive female strategy to avoid inbreeding (or outbreeding) depression when paired with a highly related (or unrelated) social mate (Kempenaers 2007). Others have argued that, because relatives share many genes identical by descent, females might increase their own inclusive fitness by providing additional breeding opportunities to genetically related males (Waser et al. 1986; Kokko & Ots 2006). Thus, in the absence of significant inbreeding depression, pursuing EPP with relatives should be favoured by kin selection, although there exist few unambiguous empirical examples of such preferences in the literature. In this issue of Molecular Ecology, Wang & Lu (2011) present an analysis of mating patterns with respect to genetic relatedness of social and extra-pair partners in the ground tit (Parus humilis), a facultative cooperative breeder in which socially monogamous pairs occasionally form cooperative groups with unpaired helper males (Fig. 1). Consistent with the predictions of the kin-selection hypothesis, females in both bi-parental and cooperative groups preferentially engaged in extra-pair matings with relatives, irrespective of relatedness to their social mates, and while suffering no apparent costs of inbreeding depression in their progeny. These findings have several exciting implications for our understanding of avian mating system diversity and the evolution of cooperative breeding.

Figure 1.

 In both (a) bi-parental and (b) cooperative breeding groups of ground tits (Parus humilis), extra-pair mates were more closely related to one another than were social mates, or than would be expected by chance. Photograph credits Xin Lu.

The suggestion that kin selection might influence the evolution of avian mating behaviours is certainly not a new idea (Hamilton 1964), and notions of inclusive fitness have been invoked to explain diverse phenomena including the participation by subordinate males in cooperative courtship displays to assist primary males in attracting mates (e.g. Krakauer 2005) and caring of offspring by nonbreeding helpers (Cockburn 1998). Likewise, several authors (e.g. Waser et al. 1986) have argued that mating with relatives is likely to be favoured by kin selection, although there is a conspicuous lack of evidence for such mate preferences, arguably due to the common presumption of severe inbreeding depression in wild populations. Indeed, cases in which high levels of inbreeding are observed (e.g. Keller & Arcese 1998) are frequently attributed to constraints on mating opportunities or an inability to recognize kin. However, theory suggests that mating with relatives should be expected under a range of conditions, provided that the benefit of inclusive fitness outweighs any costs of inbreeding depression (Kokko & Ots 2006), and mating with kin does not limit subsequent mating opportunities (Waser et al. 1986). In most cases, EPP carries few costs and does not preclude males from obtaining other mates (Griffith et al. 2002), and thus, female acceptance of kin as extra-pair mates should be favoured. To date, only one study (Kleven et al. 2005) has reported the patterns of EPP that are consistent with the kin-selection hypothesis.

Species with cooperative breeding are often found to possess the capacity for discriminating between relatives and nonrelatives (Komdeur et al. 2008), making them particularly attractive candidates for testing predictions of the kin-selection hypothesis for EPP. In their study of ground tits breeding on the north-eastern Tibetan plateau (Fig. 2), Wang & Lu (2011) utilized a combination of detailed field observations and molecular paternity analysis to reveal EPP in 36% of bi-parental and 90% of cooperative breeding groups. Interestingly, social mates appeared to pair randomly with respect to relatedness (as estimated from microsatellite genotypes), resulting in matings that ranged from outbred to highly inbred. However, the occurrence of EPP across nests was independent from the relatedness of social mates—an important finding in that it suggests females were not pursuing EPP as a means to compensate for pairing with a highly related or unrelated social mate. Furthermore, extra-pair mates were significantly more related than social mates and that which would be expected under random mating. Relatedness of these inbred extra-pair mates ranged from = 0.14–0.29, indicating that females pursued EPP with males of intermediate kinship (i.e. second or lesser degree relatives). Across the study, Wang & Lu (2011) estimated that, by engaging in extra-pair matings with male relatives, female ground tits produced, on average, an additional 0.39 offspring via inclusive fitness. Moreover, the authors found no evidence that such preferences carried fitness costs from inbreeding depression, as offspring sired by extra-pair males were found to have similar heterozygosity and physiological condition as their within-pair half siblings.

Figure 2.

 Typical ground tit habitat in alpine steppe of the north-eastern Tibetan plateau. Kin selection may help promote social cohesion in the face of high interannual population turnover observed in the study site. Photograph credit Xin Lu.

Although numerous studies have reported nonrandom patterns of mate choice with respect to male genotype (Kempenaers 2007), the mechanisms by which females evaluate male genetic quality are often unknown. In some cases, these mechanisms may be passive in the sense that females do not assess indicators of genetic quality per se (Komdeur et al. 2008). Instead, when the genetic composition of potential mates is reliably structured across the breeding season (Oh & Badyaev 2006) or space (Foerster et al. 2003) as a consequence of ecological or demographic processes, females may be able to find high-quality males by sampling in the appropriate time or place. To address this possibility, Wang & Lu (2011) examined the relatedness of potential EPP sires within the geographic vicinity of each female that engaged in extra-pair mating. They found that genetically dissimilar males occurred more often than relatives, suggesting that the observed patterns of inbred extra-pair matings resulted from an active female preference for particular males as opposed to simply an artefact of spatial clustering of kin in the study site.

In addition to providing new evidence for an adaptive function of EPP, the study by Wang & Lu (2011) lends insight into the role of kin selection in the evolution of cooperative breeding (Cockburn 1998). Unlike breeding systems where philopatric sons remain as helpers at their parents’ nests, cooperative groups in ground tits form by the addition of failed male breeders to established pairs. Previous work has shown that these helpers sire the majority of extra-pair offspring within the group and also tend to be relatives of the primary male (Du & Lu 2009). From the perspectives of helper males, such kin structure should yield inclusive fitness benefits from cooperation. At the same time, primary males benefit directly from help at the nest, while inclusive fitness from extra-pair paternity gained through related helpers may mitigate the costs of cuckoldry. The present study by Wang & Lu (2011) expands further on this finding in that it shows primary females in cooperative groups are more related to male helpers than expected by chance. Thus, by breeding cooperatively, females stand to benefit not only from assistance with parental care, but also via inclusive fitness from extra-pair offspring sired by related helper males. On a broader scale, the inclusive fitness benefits mediated through kin-directed extra-pair mating may be especially important in maintaining cooperation in this species, which experiences high interannual population turnover that might otherwise preclude the types of long-term affiliations that facilitate cooperation in other taxa (Cockburn 1998). Indeed, the results of Wang & Lu (2011) suggest that affiliations between females in bi-parental groups and their extra-pair mates, which are typically males in other socially monogamous pairings, may facilitate the formation of cooperative groups by the recruitment of these males as helpers in the event that the extra-pair male’s own nest fails.

Overall, the results presented by Wang & Lu (2011) provide support for a hypothesis that has gone largely unappreciated by current theory pertaining to extra-pair mating behaviour (e.g. Griffith et al. 2002). Whether the inclusive fitness benefits demonstrated might explain extra-pair mating in other socially monogamous bird species is an important consideration for future studies. Additionally, this work raises several questions that warrant further investigation. First, given the apparent precision of extra-pair mate choice based on relatedness, how do females discriminate among kin and non-kin? While kin recognition is commonly reported in cooperative breeding passerines (Komdeur et al. 2008), Wang & Lu (2011) concede that the phenotypic cues by which female ground tits accurately identify second and lesser degree male relatives are at present unknown. Second, in view of the inclusive fitness benefit of inbreeding, why do females mate randomly with respect to relatedness to their social mate? For the population of ground tits they studied, Wang & Lu (2011) argue that high interannual population turnover might limit availability of kin for social mates early in the breeding season. Alternatively, if social mates provide distinct types of fitness benefits (e.g. parental care), females may employ different criteria when choosing social versus extra-pair mates. Thus, future studies are likely to benefit from a whole-organism perspective that incorporates direct, indirect and inclusive fitness components of female life history (Badyaev & Qvarnström 2002).

K.P.O. is interested in the evolutionary ecology of sexual traits in natural populations.