Adult long-tailed tit showing rings that allow individual recognition in the field. (Online version in colour.) 

Context in source publication

Context 1

... the contention that kin discrimination is a widespread trait among cooperatively breeding vertebrates [25,26]. These studies all imply a role for kin selection, but rather few studies have evaluated the relative importance of direct and indirect benefits in the evolution or maintenance of helping. Among the first attempts to do so were studies by Vehrencamp [27], Woolfenden & Fitzpatrick [28], Koenig & Mumme [29] and Russell & Rowley [30], reviewed by Emlen [31], that estimated the ‘index of kin selection’ [27]—the ratio of indirect fitness to inclusive fitness. However, the estimation of fitness components in these early studies often required assumptions about relatedness that, in retrospect, are invalid, and, in species with strongly age-structured life histories, assessment of alternative options for individuals is problematic [32,33]. More recent studies have attempted to quantify the direct and indirect components of inclusive fitness informed by genetic analysis, e.g. in Seychelles warblers Acrocephalus sechellensis [34] and long-tailed tits Aegithalos caudatus [35], and in the latter case, following Woolfenden & Fitzpatrick [28], fitness estimates were based on long-term measure of individuals’ lifetime reproductive success. Despite the important advances made by these studies, most of which have strongly supported kin selection as a key process in avian social evolution, very few studies have attempted to directly test a key prediction of inclusive fitness theory, Hamilton’s rule. Hamilton [1] argued that heritable social traits will be selected for when the cost of the social action ( c ) is outweighed by the indirect benefit, which is the product of the benefit to the recipient ( b ), weighted by the relatedness ( r ) of the recipient to the actor: rb . c . The scarcity of direct tests of Hamilton’s rule does not only apply to social birds but is a more general problem. In a recent review, Bourke [36] identified 12 studies across all taxa that provide genetic and demographic data from natural populations that allow an explicit test of Hamilton’s rule to be conducted, of which three involved vertebrates and just one a cooperatively breeding bird [37]. The aforementioned studies that quantified Vehrencamp’s [27] index of kin selection, while offering important insights into the relative importance of kin selection, do not test Hamilton’s rule per se . This relative paucity of evidence in support of a key prediction of inclusive fitness theory has led some authors to question the validity and utility of the theory itself [38]. In this paper, we use data collected from a long-term field study of a cooperative breeder, the long-tailed tit ( figure 1), to test Hamilton’s rule. Long-tailed tits have several advantages over most cooperative species for this analysis, the most important being the relative simplicity of their cooperative breeding system, in which all helpers are failed breeders that redirect their care to help feed nestlings belonging to other pairs. Furthermore, they are very short lived compared with most cooperative species, facilitating measurement of the costs and benefits of alternative behaviours and allowing the rapid accumulation of complete life histories for the estimation of lifetime reproductive success [35]. We first estimate the parameters r , b and c , and then test whether Hamilton’s condition for the evolution of apparently altruistic helping behaviour is satisfied in this species. The cooperative breeding system of long-tailed tits has been described in detail elsewhere [39,40], so here we simply outline its main features. Long-tailed tits spend the non-breeding season in flocks that occupy large non-exclusive ranges that typically comprise 6–16 birds, including members of one or more nuclear families and a number of unrelated immigrants who disperse between flocks during the autumn and winter. In early spring, flocks start to break up and pairs form. There is an approximately equal adult sex ratio, and at the start of the breeding season all birds attempt to breed independently in socially monogamous pairs that occupy undefended, non- exclusive ranges. Each pair builds their own nest, an intricate domed structure that is normally sited in vegetation within 1–3 m of the ground. Clutch size is typically 9–11 eggs (mode 1⁄4 10); the female incubates alone for 13–16 days, com- mencing incubation on the day of clutch completion. Chicks hatch synchronously and are fed in the nest for 16–17 days until fledging. Families may merge soon after fledging, and dispersal between flocks commences once juveniles are independent and continues throughout the non-breeding season. Long-tailed tits are single-brooded, but they suffer a high nest-failure rate, caused mainly by predators (72% of all nests). Early in the season, failed breeders attempt to breed again, but if failure occurs after late April or early May, pairs abandon breeding for that year and a proportion of these failed breeders (especially males; 85% of all helpers) become helpers at the nest of another pair, assisting them by feeding their nestlings and fledglings [41]. As a result of the high nest-failure rate and resulting failed breeders, about half of all successful broods have helpers and helped nests have a mean of 1.8 helpers [42]. We have studied a population of long-tailed tits in the Rivelin Valley, Sheffield, UK (53 8 23 0 N, 1 8 34 0 W) since 1994. The population varies in size, but averages about 100 breeding adults during the breeding season, of which more than 95% are ringed each year with unique combinations of colour rings. We attempt to find all nests and closely monitor them, recording the timing of breeding, clutch size, hatch date (day 0), and the identity and provisioning rates of carers on alternate days from day 2 until fledging or nest failure. In the event of nest failure, we intensively search the study area for new attempts. Nestlings are ringed, weighed and measured on day 11 of the nestling period, and a small blood sample taken by brachial venipuncture (under UK Home Office Licence); blood samples are also taken from all adults at the time of first capture. DNA is extracted from blood samples and all sampled birds are genotyped at 19 microsatellite loci and sexed using two independent sex markers [43,44]. Importantly, the study has followed the same protocols since its inception, with a similar intensity of fieldwork in each year, with the exception of 2001, when access to the study site was severely constrained by restrictions imposed following an outbreak of foot-and- mouth disease; data from 2001 are therefore excluded from most analyses. In our analyses, we consider the costs and benefits of helping for an average helper, regardless of their sex. However, to determine the effect of helpers on their own and recipients’ fitness, we estimate the marginal effect of an individual helper on the current and future productivity of male recruits, as previously [41]. We focus on male recruitment because most juvenile females disperse out of the study population in their first winter, while most males are philopatric, so the local recruitment rate of all fledglings from a given brood will be a function of brood sex ratio. By measuring the recruitment rate of male offspring only (determined genetically), we can be reasonably confident that we have detected all survi- vors, and we reduce the confounding effect of dispersal on survival estimates. In previous studies of long-tailed tits [39,45] and from the early days of this study [46], it was apparent that helpers typically redirect their care towards relatives. This pattern does not ...