- Top of page
In many cooperatively breeding birds, philopatric young form queues on the natal territory, which they either inherit, or use as a staging post to gain vacancies in neighbouring territories. The outcomes of queuing can be quite diverse. In many species queues are stable as subordinates do not contest their position in the queue and individuals inherit their natal territory in a predictable order (Wiley & Rabenold 1984). In other species, power struggles can erupt once the dominant bird dies (Heinsohn et al. 2000). However, in a few species territorial inheritance appears not to occur, so queuing only provides individuals with access to vacancies on neighbouring territories (Legge & Cockburn 2000; Ekman et al. 2001; Kokko & Ekman 2002). This variability suggests that the costs and benefits of queuing also vary; yet despite considerable early interest (Wiley & Rabenold 1984; Zack & Rabenold 1989; Zack 1990), there has been little recent empirical attention to this problem.
The widespread observation that individuals do not contest their position in queues has led to growing theoretical interest in the question of why reproductive queues should be stable. There are four broad classes of explanations. First, queue members may collude to punish any individual that challenges for a higher rank, making such challenges impossibly costly (Maynard Smith 1983; Crespi & Ragsdale 2000). Second, the costs of fighting may be so great that the costs of fighting for dominance may exceed the benefits gained from promotion (Maynard Smith 1983). Third, the rank in the queue could be correlated with fighting ability, so there is no prospect of success for subordinates challenging for promotion (Maynard Smith 1983). Finally, subordinates may obtain fitness benefits while queuing.
The first two explanations are unlikely to be generally applicable. In the first case, the argument that many individuals can combine to punish challengers does not apply where the queue contains just one subordinate. In the second case, while arbitrary conventions may be feasible but difficult to assess empirically (Kokko, Lopez-Sepulcre & Morrell 2006), they may be difficult to apply when more than two individuals compete in the queue (Beacham 2003).
By contrast, differences in fighting ability with rank could easily arise because greater age could be correlated with greater ability or motivation (Krebs 1982) to defend ownership of the territory. However, this explanation has some limitation, as it deals less well with the case where same-aged individuals such as nest-mates are participants in the queue. There is also a growing literature associated with the benefits that subordinates may obtain while waiting to achieve the dominant position. Benefits can be direct if subordinates obtain some parentage, or indirect, if the subordinate obtains fitness by enhancing the production, quality or survival of relatives. Benefits can also be immediate, or deferred, in the form of territory inheritance or enhancement of the survival of relatives. Because it is a common benefit, considerable attention has been paid to whether eventual inheritance (Kokko & Johnstone 1999; Ragsdale 1999; Cant & Field 2001; Cant & English 2006; Cant, Llop & Field 2006), or privileged access to nearby vacancies (Kokko & Ekman 2002; Ridley, Yu & Sutherland 2005) can stabilize the benefits of adopting a subordinate position. In the latter case, membership of a queue allows repeated interactions with neighbours, and this gives neighbours, including subordinates, a benefit relative to floaters and intruders in contests over space. However, the stabilizing role of future inheritance is likely to depend on the relative mortality schedules of dominants and subordinates (Tsuji & Tsuji 2005; Mesterton-Gibbons, Hardy & Field 2006).
Molecular studies of the extent to which subordinates obtain immediate reproductive benefits in cooperatively breeding birds are burgeoning, but have revealed bewildering diversity (Cockburn 2004). These studies reveal one reasonably common pattern among species with natal philopatry; namely that helpers may experience two radically different circumstances before they become dominant. In the first of these, where they live on the territory with their opposite-sexed parent, mating may be constrained by avoidance of nuclear family incest (Koenig & Haydock 2004). Immediate fitness benefits available during this phase are therefore restricted to augmenting the reproductive success of relatives, acquiring reproductive skills, and competing for any extra-group parentage. By contrast, if the opposite-sex parent dies, subordinates often obtain immediate fitness benefits via within-group reproduction (e.g. Rabenold et al. 1990; Haydock, Parker & Rabenold 1996; Lundy, Parker & Zahavi 1998; Cockburn 2004).
This dichotomy has not been explored theoretically, but has some profound implications. For example, Magrath (2001) has suggested that help provided by supernumerary white-browed scrubwren Sericornis frontalis principally benefits inexperienced yearling females. However, because all females disperse in this species, these subordinates are unlikely to be related to the yearling females they assist, and in general, subordinates in this species are more likely to provision unrelated offspring (Magrath & Whittingham 1997). By contrast (Ekman, Bylin & Tegelstrom 2000; Ekman et al. 2004; Ekman 2006) have argued that offspring living with a parent can potentially gain ongoing advantages from nepotistic favouritism allowing access to food and preferred feeding sites. If this is generally true, the probability of surviving to inherit the dominant position may depend on whether the subordinates are living with their mother or an unrelated female.
Here we document patterns of survival and transition probabilities between reproductive phases in the superb fairy-wren Malurus cyaneus (Ellis, 1782), a species in which males form queues for the dominant position on the territory. Dominant status is known to be the preferred state in this species, as experimental studies show that subordinate males will always disperse to join an unmated female on a neighbouring territory (Pruett-Jones & Lewis 1990). However, kinship is unlikely to be an important stabilizing force, as extra-group paternity is extremely common (Mulder et al. 1994a), and there is no evidence that the presence of helpers aids within-group productivity (Cockburn et al. 2008). However, the presence of helpers does enhance the life span of the breeding female (Cockburn et al. 2008), extending the period where incest avoidance governs within-group mating. Here we show that queues are extremely stable, and evaluate explanations for the stability of queuing by examining demography of subordinates in relation to dominant male survival and relatedness to the female.
- Top of page
Male Malurus cyaneus form stable queues for dominant status, and the majority of birds achieve dominant status and die on their natal territory. Inheritance of the natal territory is the main path to dominant status, though dispersal to neighbouring vacancies and territorial fission are also common. While increases in ability and motivation among older birds may contribute, direct reproductive benefits while in the queue may also explain the stability of queues. We found no evidence that nepotism facilitates survival of offspring, or that the survival of dominants and subordinates was affected by fundamentally different processes.
Male Malurus cyaneus are extremely philopatric, with 60% of males gaining their first breeding vacancy on their natal territory. So far as we aware, this is the highest proportion reported for cooperatively breeding birds where just a single female usually nests on the territory, though more than 50% of males also achieve dominant status on their natal territory in two other species of Malurus (Russell & Rowley 1993, 2000) and Campylorhynchus nuchalis (Yáber & Rabenold, 2002).
The queues formed by male M. cyaneus are highly stable, with overturn of the hierarchy only occurring in two exceptional cases (of 596 males). This stability is unlikely to result solely from the prohibitive costs of fighting, as dominants could be overthrown by their neighbours. It is also unlikely that collusion among members of the queues to punish defectors is primarily responsible for enhancing stability, as there is often just a single subordinate, so any conflict is simplified to an interaction between the dominant and that subordinate. By contrast, asymmetries in fighting ability could contribute to the stability of queuing. First, violent contests leading to the eviction of neighbours were generally between males that were similar in age, and older males were also able to re-assert dominant status over younger birds when territories fused. Second, it is known that the males increase the time they spend in nuptial plumage, and hence in extra-group courtship display, for the first 5 years of life (Mulder & Magrath 1994b; Dunn & Cockburn 1999). Such improvement is suggestive of increased condition. However, this is unlikely to be a complete explanation, as neighbouring birds of similar or even lesser age do usurp vacancies. In addition, the competitors for inheritance are often same-aged nest-mates, yet appear to have adopted clear subordinate/dominant roles prior to the availability of the vacancy, usually allowing us to predict which nest-mate would gain the territory.
While the benefits associated with age are therefore ambiguous, some of the proposed general benefits from membership of queues apply in this species. Two-thirds of subordinates gained a dominant role before they died. Mesterton-Gibbons et al. (2006) have argued that for stabilization of the queue to occur exclusively as a consequence of future inheritance, subordinates must have lower early mortality and higher later mortality than dominants. That pattern was not supported in this case. Survival of dominants and helpers is extremely similar (Fig. 3a), and conforms closely to constant mortality hazard for the first 4 years of life (Fig. 3b), though the average time helpers spend in queues (48 weeks) is less than a quarter of that time. There has been considerable recent controversy over whether higher survival is expected in dominants or subordinates. The once widespread view that subordinates suffer more stress than dominants has not been supported by a number of studies of social mammals, which found the converse to be true (Creel 2001, 2005), though high stress does appear to cause reproductive suppression in meerkats (Young et al. 2006). For fairy-wrens, we found no difference in the survival of dominants and helpers, although mortality of both accelerated after 4 years. Increased mortality in older individuals is compatible with hypotheses of both senescence and terminal reproductive effort. However, in this case, we suspect that the pattern arises because older birds invest in extra-group courtship over prolonged periods, and suffer consequential costs, which escalate most strongly at about 4 years of age (Peters 2000; Peters et al. 2000; Peters, Astheimer & Cockburn 2001).
Surprisingly, and contrary to Ridley et al.'s (2005) hypothesis that a primary benefit of queuing is the establishment of reciprocal relationships that facilitate gaining vacancies on neighbouring territories once vacancies arise, older males were not more likely to gain dominant status by dispersal. The failure of older subordinates to win neighbouring vacancies could arise if dispersal was perceived as an inferior option to inheritance, because territories where vacancies arise are of inferior quality, as has been suggested for Campylorhynchus griseus (Haydock et al. 1996). However, experimental evidence suggests this is unlikely to be true in M. cyaneus. Pruett-Jones & Lewis (1990) created territorial vacancies by removing unassisted pairs from territories. Subordinate neighbours ignored the vacant territories but immediately dispersed when the female was released back on to her territory, suggesting that the presence of a female rather than territory quality is the prompt for dispersal.
Fairy-wrens are not a good candidate for indirect benefits associated with aiding kin, as subordinates are very frequently unrelated to the dominant male (Dunn, Cockburn & Mulder 1995), and there is no evidence that helpers bolster productivity of the dominants (Cockburn et al. 2008). Indeed, we found that the number of males assisting their mother or unrelated females during their lives was comparable (Table 2), and that the assistance of unrelated females was more prolonged (Fig. 2a).
The greater duration of subordinate status for unrelated males was not caused by different survivorship, but rather because the status of males helping their mother changed both because of the death of the mother and death of the dominant male. Nepotism towards offspring after fledging has been implicated as a founding condition for natal philopatry, and hence ultimately cooperative breeding (Ekman et al. 2000, 2004; Ekman 2006). We found no survival benefits of living with a mother rather than an unrelated female, suggesting that such nepotism does not occur in fairy-wrens.
By contrast, there is considerable evidence that immediate fitness benefits could be important. The extra-group paternity that dominates parentage in fairy-wrens is gained during pre-dawn forays by the female to the territory of an attractive sire (Double & Cockburn 2000). Subordinates of these attractive sires participate in the dawn chorus and are highly effective at reproductive parasitism, gaining 21% of all extra-group paternity (Double & Cockburn 2003). In addition, while there is a complete avoidance of nuclear family incest, subordinates living with unrelated females gain 22% of within-group paternity if they are not living with their mother (Cockburn et al. 2003). Hence, unrelated subordinates have both within- and extra-group opportunities for reproduction, while subordinates living with their mother are confined to extra-group opportunities. It is interesting that the presence of subordinates increases the survival of the breeding female (Cockburn et al. 2008), which should postpone the access to related subordinates to reproduction, and increase the probability of incestuous pairing if the subordinate is promoted.
We hope that these data illustrate some of the complexities posed by reproductive queuing in cooperatively breeding birds. We hope that our analysis will provoke similar approaches for other species, as the relevant data are obtained routinely in the long-term studies that have dominated the empirical approach to this question (Stacey & Koenig 1990), and may allow us to understand the diversity of mating systems and patterns of territorial inheritance among cooperative breeders.