This is a repository copy of warblers: complexities of the helping paradox.

White Rose Research Online URL for this paper: http://eprints.whiterose.ac.uk/119521/

Version: Publishers draft (with formatting)

Book Section: Komdeur, J., Burke, T. orcid.org/0000-0003-3848-1244, Dugdale, H. et al. (1 more author) (2016) Seychelles warblers: complexities of the helping paradox. In: Koenig, W.D. and Dickinson, J.L., (eds.) in vertebrates: studies of ecology, evolution, and behavior. Cambridge University Press , Cambridge , pp. 181-196. ISBN 9781107338357 https://doi.org/10.1017/CBO9781107338357.013

© Cambridge University Press 2016. This material has been published in Cooperative breeding in vertebrates: studies of ecology, evolution, and behavior edited by Koenig, W.D. and Dickinson, J.L.. This version is free to view and download for personal use only. Not for re-distribution, re-sale or use in derivative works.

Reuse Unless indicated otherwise, fulltext items are protected by copyright with all rights reserved. The copyright exception in section 29 of the Copyright, Designs and Patents Act 1988 allows the making of a single copy solely for the purpose of non-commercial research or private study within the limits of fair dealing. The publisher or other rights-holder may allow further reproduction and re-use of this version - refer to the White Rose Research Online record for this item. Where records identify the publisher as the copyright holder, users can verify any specific terms of use on the publisher’s website.

Takedown If you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing [email protected] including the URL of the record and the reason for the withdrawal request.

[email protected] https://eprints.whiterose.ac.uk/ 12

Seychelles warblers: complexities of the helping paradox

Jan Komdeur , Terry Burke , Hannah Dugdale , and David S. Richardson

Introduction 1992 ; Richardson et al. 2002 ; Eikenaar et al. 2008 , 2010 ). By the 1960s the last remnant population of this then h e Seychelles warbler ( sechellensis), a critically endangered species was coni ned to Cousin endemic to the Seychelles archipelago, is a Island (Figure 12.1). Subsequent conservation actions, facultatively cooperative breeder that lives either in pairs including the restoration of forest habitat and the estab- or small groups. Breeding groups normally consist of a lishment of new populations through translocations, dominant pair and one to three subordinates, although have provided unique opportunities to study the evolu- up to nine have been observed. Subordinates can be of tionary ecology of cooperative breeding in this species. either sex, and are often of spring that have delayed dis- We have followed the entire world population of the persal and remained in their natal territory (Komdeur Seychelles warbler since our study started in 1981. In

Cooperative Breeding in Vertebrates: Studies of Ecology, Evolution, and Behavior , eds W. D. Koenig and J. L. Dickinson. Published by Cambridge University Press. © Cambridge University Press 2016.

197

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 197197 99/28/2015/28/2015 3:38:203:38:20 PPMM 198 Jan Komdeur et al.

(a) (b)

Figure 12.1. (a) , with Cousine Island in the distance. Photograph by David Richardson. (b) Seychelles warblers feeding a nestling. Photograph by Danny Ellinger. See plate section for color i gure.

1997, however, we increased the intensity of work on system and the selective factors that favor switching Cousin and since then nearly all individuals, includ- between dif erent forms of cooperative and indepen- ing l edglings, have been captured, ringed, blood sam- dent breeding. pled, sexed, and monitored for breeding and status. Molecular tools have been used to assign the sex and genetic parentage of young and to determine Distribution and conservation levels of relatedness between individuals. In addi- tion, a comprehensive set of behavioral, life history, h e Seychelles warbler is a small (13–19 g) passerine and annual i tness parameters have been recorded for endemic to the Seychelles archipelago in the Indian nearly all individuals, providing important opportuni- Ocean (Saf ord and Hawkins 2013 ; Figure 12.2 ). In ties for assessing changes in social behavior. h e lack of the 1870s this species was recorded on the islands of interisland dispersal (Komdeur et al. 2004a ), combined Mahé, Marianne, Félicité, and Cousine (Oustalet 1878 ; with sampling of the entire population, provides a rare Figure 12.2). It probably also occurred historically on opportunity to monitor the survival, reproduction, and most of the Seychelles islands, which made up a sin- lifetime i tness of all individuals within the population. gle large island during the last ice age (Collar and Stuart Our long-term research program into cooperative 1985 ). h is assumption is supported by the large ef ec- breeding, hand-in-hand with conservation actions, tive population sizes estimated for the pre-bottleneck has created an experimental system in which we can Seychelles warbler population using museum samples attempt to unravel the factors that alter the cost-benei t from the 1800s (Spurgin et al. 2014 ). trade-of s of cooperative breeding. Over time, this sys- h e destruction of natural habitat for the planting tem has become proof of the power of the experimental of coconut trees (Cocos nucifera), along with the intro- methods and of the corrective value of long-term stud- duction of mammalian predators in the early 1900s, ies, where iterative examination with longer-term data resulted in the warbler’s extirpation from nearly all sheds new insights on short-term i ndings. Here we the islands where they previously occurred. Only detail various i ndings that have allowed us to uncover on Cousin Island, which remained free of predators how changing social and ecological factors inl uence (Collar and Stuart 1985 ), did the warblers survive, with the form and function of reproductive competition and only 26–50 individuals remaining between 1940 and cooperation in the Seychelles warbler. We also out- 1967 (Crook 1960 ; Loustau-Lalanne 1968 ; Spurgin et al. line how molecular genetic tools have given us a bet- 2014 ). Cousin Island was purchased by a consortium ter understanding of the species’ cooperative breeding led by BirdLife International (then the International

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 198198 99/28/2015/28/2015 3:38:223:38:22 PPMM Seychelles warblers 199

1990, respectively (29 individuals each time; Komdeur INDIA Denis 1994a ), Denis in 2004 (58 individuals; Richardson e t a l . AFRICA 2006 ), and Frégate in 2011 (59 individuals; Wright et al. EQUATOR 4°00 Seychelles 2014 ; Figure 12.2). In contrast to the food limitation Aride observed on Cousin, where a shorter breeding season Félicité was the norm, warblers on the new islands experienced Cousin Cousine higher food availability and initially bred year-round Marianne 4°30 (Komdeur et al. 1995 ). h is resulted in an initial increase in annual productivity of up to ten-fold com- Frégate pared to the population on Cousin (Komdeur 1996a ). Mahé As each population increased, productivity declined, Seychelles N presumably because of the increased density of the 5°00 population (Brouwer et al. 2009 ). Most recent popu- 40 km lation sizes are estimated at 210 warblers on Cousine 55°00 55°30 56°00 in 2007 (van de Crommenacker and Richardson 2007 ), Figure 12.2. h e location of the Seychelles in the Indian 1850 on Aride in 2003 (Orchard 2004 ), 300 on Denis in Ocean (inset) and the inner granitic islands in detail, 2013 (J. van de Woude, unpubl. data), and 80 on Frégate including Cousin (0.29 km2 ), Cousine (0.25 km2 ), Aride in 2013 (Teunissen 2013 ) ( Figure 12.4). h e world pop- (0.68 km2 ), Denis (1.42 km 2 ), Félicité (2.68 km 2 ), Frégate ulation of Seychelles warblers is now estimated at 2750 2 2 2 (2.19 km ), Marianne (0.96 km ), and Mahé (154 km ). adult birds across i ve islands and the conservation status of the Seychelles warbler has been reduced from Committee for Protection) in 1968 and designated endangered to vulnerable (IUCN 2013 ). as a nature reserve to save the Seychelles warbler from . Habitat restoration was successful and, by 1982, much of Cousin was again covered with native Natural history: Seychelles warblers forest (Bathe and Bathe 1982 ). As a consequence, the warbler population recovered and since 1982 has been Territoriality and breeding biology at a carrying capacity of about 320 adults inhabiting around 110 territories that cover all but the bare rock h e Seychelles warbler is socially monogamous, areas of Cousin ( Figure 12.3 ). h is ratio of birds to ter- and paired dominant birds often remain with the ritories also means there has been a considerable sur- same partner on the same breeding territory for life plus of (unpaired) adult birds on the island since the (Komdeur 1992 ; Richardson et al. 2007 ). h e pair bond point of saturation (Komdeur 1992 ; Brouwer et al. can last up to 14 years (S. A. Kingma, unpubl. data). 2009, 2012). Seychelles warblers are largely insectivorous, mainly h e l ight apparatus of the Seychelles warbler does gleaning invertebrate prey from the undersides of not dif er from that of other long-distance migra- leaves. As such, resource abundance, a major aspect tory Acrocephalus w a r b l e r s ( K o m d e u r e t a l . 2 0 0 4 a) . of territory quality, can be evaluated by estimating the Nonetheless, interisland dispersal is virtually nonex- number of invertebrates on the undersides of leaves istent. During the 28 years of this study only six indi- and extrapolating from this to estimate prey abun- viduals were recorded to have crossed the sea between dance according to the amount of foliage in each terri- islands. Given the low probability of any new warbler tory (Komdeur 1992 ; Brouwer et al. 2006 ). On Cousin, populations establishing by themselves, and the vul- prey abundance has been measured monthly on each nerability of a single small population, four new popu- territory during the main breeding periods (June to lations were established by translocation of birds from September) for most years since 1987; the same tech- Cousin to the islands of Aride and Cousine in 1988 and nique has been used to measure prey abundance in

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 199199 99/28/2015/28/2015 3:38:233:38:23 PPMM 200 Jan Komdeur et al.

1986 – 1990 1991 – 1995 2000 – 2008

mean insect abundance (x1000) 0 – 25 100 m 26 – 50 51 – 75

Figure 12.3. Cousin Island outline with Seychelles warbler territories drawn in and mean insect abundance present in each territory during the periods 1986–1990 (from Komdeur 1992 ), 1991–1995, and 2000–2008 (J. Komdeur, unpubl. data).

Cousin 10

Aride

N=29 Cousine 1 N=29 N=58 N=59 Population density (no. birds/ha) Denis Frégate

1982 19861990 1994 1998 2002 2006 2010 2014 Year

Figure 12.4. Temporal l uctuations in the population density (ln scale) of Seychelles warblers on Cousin (saturated), Aride (census focused on the accessible plateau area of 5.2 ha; saturated), Cousine (saturated), Denis (increasing phase), and Frégate (increasing phase) islands. Arrows with sample sizes (N birds) indicate the introduction of birds to each island. Initial densities dif er both because of the dif erent numbers of birds introduced and large dif erences in the sizes of the islands (Cousin: 0.29 km2 , Aride: 0.68 km2 , Cousine: 0.25 km2 , Denis: 1.42 km 2 , Frégate: 2.19 km 2 ). Figure updated from Brouwer et al. (2009 ).

various years on the other islands since their popu- synchronized with the periods of highest prey abun- lations were founded. Our results show that insect dance (Komdeur 1996a ) with peaks of nest building food supply is highly seasonal and that warblers time between May and August and again, to a lesser extent, their breeding such that the production of nestlings is between December and February.

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 200200 99/28/2015/28/2015 3:38:233:38:23 PPMM Seychelles warblers 201

Most breeding pairs and groups produce one clutch young or laying eggs) may help to raise the chicks, per year (Komdeur 1996a ), with clutches normally con- either by aiding in nest building (mainly females), incu- sisting of just one egg, although ~13% of nests contain bation (females), guarding the clutch (mainly males), two or three eggs (Richardson et al. 2001 ). Incubation, or provisioning (males and females) (Komdeur 1994b ; which is by the female, lasts 17–19 days, and nestlings Richardson et al. 2003a , 2003b ); we dei ne these birds are fed by both sexes for 18–20 days prior to l edging, as “helpers” in that breeding attempt. Subordinates and then for up to four months before reaching inde- that neither help nor breed successfully are dei ned as pendence. h is is an extremely long period of depen- “nonhelping subordinates.” dence; no other known passerine has a six-month h ere is no evidence of egg dumping; females (sub- period of dependence. Indeed, the mean l edging ordinate or otherwise) do not lay eggs in extra-group period for other is only about one tenth as nests. Similarly, male subordinates never gain paternity long (Bennett and Owens 2002). Nonetheless, i rst-year outside the group (Richardson et al. 2002). However, survival is relatively low: 61% of l edglings survive the extra-group paternity is common, with about 44% of i rst year compared to an annual survival of 84% in of spring fathered by dominant males from other ter- adults (Brouwer et al. 2006 ). Like many other tropical ritories (Richardson et al. 2001 ; Hadi eld et al. 2006). bird species, the Seychelles warbler is long-lived, with Populations also include a variable, and dii cult to an average life expectancy of around i ve years once dei ne, number of l oaters of both sexes that have failed they reach adulthood and a maximum recorded lifes- to i nd a breeding or subordinate position and wander pan of 17 years (Barrett et al. 2013). over the island without residing in a territory (Komdeur and Edelaar 2001 ; Eikenaar et al. 2008 ). The breeding system

Seychelles warbler territories are normally occupied The costs and benefits of by a single dominant pair-bonded male and female. cooperative breeding However, about 30% of territories contain subordinate males and/or females that are sexually mature birds In most species the inferred i tness benei ts from help- lacking a suitable independent breeding opportunity ing are substantially lower than those from breeding (Komdeur 1992 ; Richardson et al. 2002 , 2007 ). h is is independently. Consequently, cooperative behavior is demonstrated by the fact that birds transplanted to generally considered to be a suboptimal i tness strat- previously uninhabited islands immediately breed egy (Emlen 1982 ; Lessells 1991 ; Dickinson et al. 1996 ; independently rather than becoming helpers or subor- Dickinson and Hatchwell 2004 ). h erefore, in order to dinate cobreeders. Furthermore, vacancies created in understand its evolution it is important to know what the source population by the translocation of breeders determines when and where cooperative breeding are i lled immediately by subordinates from other terri- behavior occurs. With this in mind, the route to coop- tories (see “Indirect i tness benei ts to subordinates”). erative breeding may best be viewed as a two-step Female subordinates normally remain on their process: i rst, the decision by an individual to forgo natal territory, while ~25% of male subordinates independent breeding and become a subordinate in move to a new territory to take up a subordinate posi- a group, and second, the decision by a subordinate to tion (Richardson et al. 2002). In any given breeding help the dominants to raise of spring (Emlen 1982 ). h e attempt, approximately 44% of subordinate females lay i rst step is usually attributed to the existence of eco- eggs within the dominant pair’s nest, while only 15% logical constraints on independent breeding, such as a of subordinate males sire of spring within the group shortage of breeding territories or mates (the “ecolog- (Richardson et al. 2001 ); we dei ne such birds as “sub- ical constraints” hypothesis; Emlen 1982 , 1991 ), com- ordinate cobreeders. ” h e remaining subordinates bined with the benei ts of remaining in the natal group that have not successfully reproduced (either by siring (the “benei ts of philopatry” hypothesis; Brown 1978;

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 201201 99/28/2015/28/2015 3:38:243:38:24 PPMM 202 Jan Komdeur et al.

is a positive correlation between the length of time for 60 which independent of spring delay dispersal and the 50 natal territory quality, females ended up delaying dis- persal longer than males. h is sex-bias in dispersal 40 behavior disappeared later in the study, however, 30 apparently as a result of females starting to disperse earlier, rather than males delaying dispersal longer 20 (Eikenaar et al. 2007 , 2010 ). 10 We suspect that temporal changes in demography,

% Territories with subordinates most likely higher reproductive output per territory 0 0 20 40 60 80 100 due to improved and less variable habitat quality, % Saturation rather than dif erences in local density or social condi- tions, caused the observed changes. In particular, the Figure 12.5. h e percentage of territories with subordinates in change in male dispersal behavior coincided with a relation to habitat saturation in a given year for the Seychelles change in spatial food resource availability. Early on, warblers on Cousin Island between 1959 and 2013. Habitat mean insect numbers were highest in the center of saturation was measured as the number of territorie s present in that year divided by the maximum number of territories recorded the island and decreased gradually toward the coast in a year (N = 121). Figure amended from Komdeur et al. ( 2008 ). (Figure 12.3), but with the restoration of the coastal vegetation insect numbers there increased signii - cantly (Komdeur and Pels 2005 ). Consequently, the Stacey and Ligon 1987 , 1991 ; Ekman et al. 2004 ; Covas proportion of high-quality territories on the island and Griesser 2007 ). h e second step is attributable to increased (van de Crommenacker et al. 2011 ), even a variety of potential i tness benei ts, both indirect though the total number of territories, and Seychelles and direct (Stacey and Koenig 1990 ; Dickinson and warblers, has remained similar (Brouwer et al. 2009 ). Hatchwell 2004 ). Given this change in food abundance across a more or In the Seychelles warbler, cooperative breeding was less stable population and the concomitant increase in i rst observed in 1973, when the population of adult the quality and reproductive output of territories over birds started to be greater than the number of breed- the years (Komdeur and Pels 2005 ), we suspect that ing positions available on territories. At this point, food-related i tness consequences drive the dispersal most, but not all, suitable breeding habitat on Cousin behavior of males, although further work is needed to was occupied (Komdeur 1992). Interestingly, however, coni rm this. cooperative breeding occurred well before the island was completely saturated (Figure 12.5 ), indicating that factors other than habitat saturation per se also inl u- Helping behavior ence cooperative breeding. Helping is not an inevitable consequence of group l i v - ing. Both within and among cooperatively breeding Group living species, considerable variation exists in whether, and to what extent, individuals help. Furthermore, helping in Early on in our study, young independent males dis- many species is biased with respect to the subordinate persed more often and earlier from their natal terri- sex. In most cooperatively breeding birds, male-biased tories than young independent females (Komdeur delayed dispersal and helping is the norm (Cockburn 1996b ). h is was thought to be because females are 1 9 9 8; B e r g e t a l . 2 0 0 9; C o r n w a l l i s e t a l . 2 0 0 9) . S e y c h e l l e s usually born on higher-quality territories than males warblers were originally thought to be an exception (Komdeur et al. 1997, see “”), and as there in that females were more likely to delay dispersal

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 202202 99/28/2015/28/2015 3:38:243:38:24 PPMM Seychelles warblers 203

and become subordinate helpers/cobreeders than Since Seychelles warblers can breed independently males, with 88% of subordinates being female helpers/ in their i rst year, why do some individuals become cobreeders in 1986–1990 (Komdeur 1996b ). As men- subordinates? Individuals that remain as subordinates tioned above, this bias toward helpers being females in high-quality territories and later breed there have decreased over time to 68% in later years (Richardson higher lifetime i tness than those that disperse from et al. 2002 ) and subsequently disappeared altogether high-quality territories at one year of age and breed in (Eikenaar et al. 2010 ). lower-quality territories. Breeding pairs without help- Additionally, a small proportion of birds (14% of ers on high-quality territories produce on average 0.85 females and 3% of males) became subordinates, usu- l edglings each year compared to only 0.19 l edglings ally for the second time, after being deposed as domi- on low-quality territories. Furthermore, a helper on a nant breeders. Many of these (68% of the females) were high-quality territory can increase the dominant pair’s observed to actively help kin and are therefore termed reproductive success by 0.77 l edglings from 0.85 to “grandparent helpers ” (Richardson et al. 2007 ). h ese 1.62 l edglings per year, about four times the output observations suggest that helping is not a stable trait, from the low-quality territories. h is increase, and the but rather a plastic response to local conditions. h e fact that subordinates are normally helping their par- key, then, is to assess and understand how dif erent ents, means that subordinates increase their own indi- costs and benei ts, to both the helpers and the helped, rect benei ts by remaining as helpers (Komdeur 1992 ). interact to determine this variation. At the time these results were published we assumed that all female helpers were nonbreeding helpers. Later work indicated that these "helpers" included both true Indirect fitness benefits to subordinates nonbreeding helpers and subordinate cobreeders For helpers to gain indirect bene i ts their activities must (Richardson et al. 2001 ). To the extent that these birds result in a net benei t to those they help (Hamilton were actually cobreeders, they would have gained even 1964 ). In the Seychelles warbler, the presence of help- greater i tness benei ts by remaining with their parents. ers on high quality territories with high prey abun- We were able to experimentally verify the importance dance increases the number of young produced on a of territory quality to cooperative breeding by creating territory: the removal of helpers, from groups with one new breeding opportunities when removing breed- helper on high-quality territories, resulted in lower ing adults for translocation to new islands. Initially, in reproductive success for the breeding pair compared the source population, the breeding vacancies created with control pairs (Komdeur 1994b ). A later study dem- were i lled by subordinates from territories of equiv- onstrated that the total amount of provisioning to nest- alent or poorer quality, never by subordinates from lings signii cantly af ects l edging success and i rst-year superior territories, for which helping remained a bet- survival. Importantly, this improvement was corre- ter option than breeding (Komdeur 1992). On the new lated with the number of birds actually helping, not islands, initially all translocated individuals bred as with the total number of birds in the territory includ- independent pairs. However, after a few years, when all ing nonhelping subordinates (Brouwer et al. 2012 ). the high-quality areas were occupied, individuals again Furthermore, cross-fostering of nestlings between ter- chose to help on high-quality territories rather than to ritories coni rmed that it is the number of helpers in the breed independently on low-quality territories. h ese territory where they are reared that determines the sur- results support both the benei ts of philopatry (Stacey vival of of spring, thus ruling out the possibility that of - and Ligon 1987 , 1991 ) and the ecological constraints spring quality is confounding the helper ef ect. Overall, hypotheses (Emlen 1982 , 1991 ), which dif er only in the the number of l edglings produced on a territory per emphasis they place on the costs of leaving versus the breeding attempt increases by 18% for each helper pre- benei ts of staying (Komdeur 2003 ). sent, even after excluding direct subordinate parentage Cooperatively breeding species are often long-lived, (Richardson et al. 2002 ). have stable and strong natal philopatry, and a relatively

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 203203 99/28/2015/28/2015 3:38:243:38:24 PPMM 204 Jan Komdeur et al.

long period of extended parental care during which 12 they associate with family members (Stacey and Koenig 10 1990; Emlen 1991; Arnold and Owens 1998 ; Hatchwell and Komdeur 2000; Cornwallis et al. 2009 ). Most sub- 8 ordinates become helpers after delaying dispersal from 6 their natal territory. In such situations, conditions exist that may assist subordinates to gain indirect i t- Feeds (h) 4 ness benei ts through helping: i rst, since parents are 2 long-lived, helpers on their natal territories will be car- ing for subsequent related of spring, and second, the 0 extended care period may allow subordinates (“ego”) –0.4 –0.2 0.0 0.2 0.4 0.6 to develop a simple rule: help any individual (“ b”) that Mean subordinate-nestling relatedness fed ego when ego was a nestling to raise b’s subsequent Figure 12.6. Provisioning rates by subordinate female (i lled broods, as b is likely to be ego’s parents. However, in circles) and male (empty circles) nonparents in relation groups where individuals dif er in how related they are to relatedness between the subordinate and the nestling. to each other and to the brood because of high turnover h ere is a positive relationship between provisioning and of breeders or ini delity, direct kin discrimination will subordinate–nestling relatedness for female subordinates (solid line; F = 4.82, P = 0.05, R 2 = 0.31) but not for male be required if potential helpers are to maximize their 1,11 subordinates (broken line; F = 0.04, P = 0.86, R 2 = 0.00). indirect benei ts. 1,8 From Richardson et al. ( 2003a). In the Seychelles warbler, males more often become subordinates on nonnatal territories than females and so subordinate males are no more related to the groups’ from outside the group are responsible for siring about nestlings than one would expect by chance (Richardson 40% of of spring (Richardson et al. 2001 ; Hadi eld et al. et al. 2002). Female subordinates are, therefore, more 2006 ). h is extra-group paternity dramatically reduces related to the nestlings than male subordinates. h us, the degree of relatedness between subordinates and the kin-selected benei ts of helping are more impor- the of spring they may help, as only 36% share the same tant for female subordinates. Even within the sexes, father, even if the same male remains dominant in the however, levels of relatedness and the amount of help- territory across their birth years. Second, subordinate ing vary substantially within and among subordinates. females commonly lay an egg within the dominant’s Female subordinates that gain parentage in the nest nest (44% successfully produce of spring in a breed- appear to assess this fact accurately and always provi- ing season; Richardson et al. 2002 ). h us, the of spring, sion (Richardson et al. 2003a, 2003b ). For nonparent including those now acting as subordinates, in a terri- female subordinates, provisioning is positively corre- tory with multiple females may have dif erent mothers, lated with their genetic relatedness to the nestlings and both within and between breeding seasons. h e con- thus their potential indirect i tness gain (Figure 12.6 ). sequently reduced relatedness between subordinates In contrast, male subordinates rarely obtain pater- and the helped nestlings (mean r = 0.16; Richardson nity in the brood that they help (Richardson et al. 2 001) et al. 2003b ) means that the indirect benei ts of help- and are usually less related to the nestlings. h ese males ing are relatively small, and substantially less than was provision considerably less than subordinate females expected before we were able to assess relatedness and their contribution is not inl uenced by their relat- directly using molecular techniques. edness to the nestling (Richardson et al. 2003a, 2003b ). Surprisingly, despite these complications, subor- Relatedness between subordinates that remain in dinate females appear to possess a mechanism for their natal territory and nestlings is lower than one inferring their kinship to nestlings and direct help would theoretically expect for i rst-order relatives toward those to which they are related. With its long (mean r = 0.13). h is occurs for two reasons: i rst, males period of parental care, kin discrimination is based

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 204204 99/28/2015/28/2015 3:38:253:38:25 PPMM Seychelles warblers 205

her relatedness to the nestling. For nestlings that later n = 4 967 10 became subordinates, their decision to help was asso- NS p = 0.003 ciated only with the continued presence of the domi- nant female (their putative mother) irrespective of 8 whether the subordinate was cross-fostered, and there- fore regardless of their actual genetic relatedness to that female (Figure 12.8 ). 6 It is clear, however, that the cue just outlined will only work for subordinates that were raised in a terri- tory without female helpers present. It remains to be Feeds per hour 4 investigated whether and how subordinates that were raised in groups of multiple females discriminate kin. For example, if mothers provide more food than help- 2 ers, then the subordinates might base their decision to help on the continued presence of the female that fed them the most. Alternatively, since both the breeding 0 Primary male Primary female and helping females within a group are often related (Richardson et al. 2002 ), the subordinate may help to Figure 12.7. Mean provisioning (± S.E.) by subordinate female nonparents in relation to the presence (light gray bars) feed the nestling if either potential mother remains or absence (dark gray bars) of their putative parents (primary present, as this will still indicate subordinate–nestling relatedness, although to a lesser degree. males t11 = 0.02; females t 11 = 3.71). n denotes sample size. From Richardson et al. ( 2003a ). Even after taking into account subordinate–nestling relatedness, considerable variation in the extent of helping behavior remains unexplained (Richardson on association and is achieved through learning the et al. 2002 ). To gain a full and accurate picture of the identity of parents. h e subordinate female’s provi- role of indirect benei ts in helping, the costs of help- sioning of the nestlings is best predicted by the con- ing, and not just the benei ts, need to be considered. tinued presence of the dominant female – her putative In the Seychelles warbler, helping appears to be costly mother – that raised her. Because the putative mother’s and condition dependent. Female subordinates that continued presence reliably indicates a subordinate’s help have lower body mass at the end of the season relatedness to the nestling (Richardson et al. 2003a , than female subordinates that do not help (van de 2003b ), the use of this cue is ef ective in maximizing Crommenacker et al. 2011 ). Consequently, body con- the indirect benei t that subordinates gain from their dition has a positive ef ect on an individual’s decision ef orts ( Figure 12.7 ). to help: only those individuals that are in good condi- In contrast, the continued presence of the dominant tion prior to the breeding season will help. Results such male that raised the subordinate is not used to deter- as these in the Seychelles warbler and in other species mine when to help. From an evolutionary perspective such as meerkats ( Suricata suricatta ; Russell et al. 2003 ; this makes sense because the high frequency of female Chapter 17), suggest that helpers may be only using ini delity means that the continued presence of the energy that is surplus to their requirements. If this is same dominant male does not predict reliably that sub- true, are there any actual costs of helping? Unless the ordinates are related to the nestling. proximate costs lead to a reduction in future survival Cross-fostering experiments focusing on breeding and i tness, there may not be any ultimate cost of help- pairs without female helpers have coni rmed that the ing. So, by rei ning their helping decisions based on subordinate’s decision to help is based on the identity their condition, helpers may be able to obtain an indi- of the female parent, rather than a direct assessment of rect benei t at little or no long-term cost.

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 205205 99/28/2015/28/2015 3:38:253:38:25 PPMM 206 Jan Komdeur et al.

0.7 18 0.6

n = 10 0.5 0.5 13 0.4 0.4

0.3 0.3

0.2 12 0.2 Fraction helping Fraction 0.1 2 0.1 female relatedness female 43Subordinate-primary 0.0 0.0

–0.1 2

Primary female: Present Absent Present Absent Subordinate female: Control Cross-fostered

Figure 12.8. h e fraction (gray bars) of control and cross-fostered female subordinates that helped to feed nestlings during their second year of life in relation to the continued presence of the primary females and primary foster females ( χ 2 = 1.3, d.f. = 1, P = 0.25), and in relation to the new primary females that replace the primary (foster) females that raised the subordinates (absent). Only those subordinates that could have been helpers, that is, where nests with young were present in their natal territory, were included ( N = 38 individuals). Circles indicate mean relatedness of subordinates (± S.E.) to the primary females

and primary foster females that raised the subordinates and were still present on the territory (present; t20 = 3.48, P = 0.002) or to new primary females that replaced the primary (foster) females that raised the subordinates (absent). n denotes the number of subordinates. From Komdeur et al. (2004b ).

Direct fitness benefits of helping by laying an egg in the dominant bird’s nest (Richardson It has been suggested that only 10% of the variation in et al. 2001 ). In contrast, male subordinates rarely gain helping behavior can be explained by indirect benei ts parentage. As a result, the direct i tness benei ts gained (Grii n and West 2002 ). h us indirect i tness benei ts through parentage by female subordinates are over alone are unlikely to fully explain this phenomenon. An three times higher than those gained by male subordi- alternative explanation is that helpers may increase their nates (Figure 12.9). Importantly, in both sexes the direct own survival and future reproduction by cooperating benei ts of helping are up to six times higher than the with others. Here we do not consider all the various pos- indirect benei ts when direct breeding benei ts were sible ways that subordinate individuals might improve included. h is suggests that direct benei ts are more their direct i tness, but instead focus on those that important in the maintenance of cooperative breeding appear to be important in the Seychelles warbler system. in the Seychelles warbler than are indirect benei ts.

Acquisition of parentage Accumulation of breeding experience

As discussed earlier, female subordinate Seychelles By helping to raise the dominant birds’ of spring, sub- warblers often gain parentage within their own group ordinates may gain breeding experience, allowing them

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 206206 99/28/2015/28/2015 3:38:253:38:25 PPMM Seychelles warblers 207

for inexperienced females. Once the previously inex- n = 40 20 p < 0.05 perienced birds l edged young themselves, they were then able to produce a second l edgling in the same 0.6 p < 0.001 time as experienced birds, suggesting that they had by then learned all the necessary skills. 0.5 In this study, we presumed that all the female help- ers were nonbreeders and that all dominant females 0.4 had bred. However, our later work indicated that 44% of helper females in territories with genotyped of spring 0.3 NS were assigned maternity, so it is likely that some of the p < 0.05 0.2 female helpers included were subordinate cobreeders. Either way, this study shows that the females that gain 0.1 breeding experience as subordinates do better when Benefits (offspring equivalents) subsequently breeding as a dominant for the i rst time 0.0 than females without breeding experience. Given that Female Male some "helping" females may have been cobreeders, Subordinates however, it does not unequivocally demonstrate that Figure 12.9. Fitness benei ts (± S.E.) of cooperative breeding "helping” females do better as a breeder as a result of gained by female and male subordinates (1997–1999). Both the experience or skills gained from their behavior. female and male subordinates gain signii cantly higher direct (pale columns) compared to indirect (dark columns) breeding benei ts (female: z = 4.22, males: z = 2.39). Direct breeding Territorial inheritance benei ts were signii cantly higher in females than in males Helping may lead to higher status within the group (z = 2.29), but there was no signii cant dif erence between the and, hence, greater success when competing for a ter- sexes in indirect benei ts ( z = 0.21). n denotes sample size. From Richardson et al. ( 2002 ). ritory after the death of its owner of the same sex as the helper (Zack 1990 ; Koenig et al. 1992 ; Balshine-Earn et al. 1998 ; Leadbeater et al. 2011 ). Alternatively, help- to be more productive when they achieve a breeding ing can result in higher productivity and thus a larger territory of their own in the future (Komdeur 1996c ; group size, which may make the group more compet- Hatchwell et al. 1999 ). In the Seychelles warbler, this itive compared to neighboring groups. If the group idea was tested by translocating similar-aged females then increases the size of its territory by outcompeting and males with dif erent breeding and helping expe- neighboring groups, the helper may eventually be able rience. On the new islands, newly formed pairs, where to take over a portion of the territory as its own breed- one partner only had helping experience and the other ing territory (Emlen 1991 ). had breeding experience, produced their i rst l edgling In the Seychelles warbler, territory inheritance is rare as fast as pairs in which both partners were experienced and not linked to helping: of the 219 monitored subor- breeders, and signii cantly faster than pairs consisting dinates, only 3.7% (i ve males and three females) inher- of an inexperienced breeder with no helping experi- ited their natal territory (Eikenaar et al. 2008 ). However, ence and an experienced breeder (Komdeur 1996c ). of the males that remained on their natal territory as Inexperienced females took more than a year to pro- subordinates, 78% budded of a small portion of the ter- duce a i rst l edging, but only four months when they ritory for themselves without having ever been a helper had helping experience. Females with either helping or (Komdeur and Edelaar 2001 ). h ose males that did help breeding experience built stronger nests that were less never budded of part of the territory and never inher- prone to be blown away by wind, and spent more time ited a breeding territory after the death or experimental incubating, resulting in higher hatching success than removal of the dominant male (Eikenaar et al. 2008).

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 207207 99/28/2015/28/2015 3:38:263:38:26 PPMM 208 Jan Komdeur et al.

Territory inheritance is therefore almost exclusively the of spring’s chances of reaching adulthood and its linked to budding , and not helping, and even then only late-life survival as an adult ( Figure 12.10 ). Importantly, accounts for about 4% of cases of territory acquisition. larger group size per se does not appear to be benei - h at it is only male subordinates that are able to bud cial to adults, as individuals living in larger groups had of a new territory may be because male subordinates lower survival probabilities than those living in small gain fewer i tness benei ts from helping than female groups (Figure 12.11). Predation of warbler eggs and subordinates ( Figure 12.9 ). Plus, if helping is costly in nestlings occurs, but adult warblers suf er virtually no terms of condition, males may do better by refraining predation pressure on Cousin Island; therefore compe- from helping and instead harboring their resources tition for food is the most probable cause of this reduc- to mount a challenge to bud of part of a territory and tion in adult survival with increasing group size. expel intruders . Overall, however, the negative ef ect of increasing group size on individual survival may be counterbal- anced by a gain in the reproductive success of subor- Pay-to-stay dinates (Richardson et al. 2002). For example, in larger h is hypothesis proposes that individuals that do not groups, female subordinates may have a higher chance help are punished and may be evicted from the territory of becoming cobreeders, although this remains to be (Gaston 1978 ; Balshine-Earn et al. 1998 ). Importantly, investigated. Finally, lower survival benei ts in larger the "resources" that the helpers are paying to get access groups may form an additional selection pressure on to are not necessarily food. For example, where subor- subordinates to help rear related of spring in order to dinates join and help in a group other than their natal gain a compensatory benei t (at least for female subor- group, the goal of helping may be to form social or breed- dinates) or to disperse and join a smaller group to of set ing relationships with nonkin (Croft et al. 2004 ). h is the negative ef ects of increasing group size as an adult . could well be one of the reasons why, in the Seychelles warbler, males – the sex that gains lower i tness benei ts from helping on the natal territory than female subordi- Sex allocation nates – often disperse to become subordinates on non- natal territories, while females normally remain on their In the Seychelles warbler, the advantage of having help- natal territory as subordinates (Richardson et al. 2002 ). ers depends on territory quality. Helpers are benei cial If this is true, the i tness benei ts to subordinate males for dominant birds on high-quality territories because on nonnatal territories should be higher than those to they improve the reproductive success of breeding males on natal territories; this remains to be tested. pairs. Unassisted dominant females on high-quality ter- ritories maximize their i tness by biasing the sex ratio of eggs to females (Figure 12.12 ) thus producing a greater Group augmentation proportion of female of spring that are likely to become Helping may result in the production of more individu- helping subordinates in subsequent years (Komdeur als, causing the group to grow larger. h is may increase 1 9 9 6 b, K o m d e u r e t al. 1997 ). When the group includes the survival chances of all individuals in the group, helping female subordinates, the dominant birds not because larger groups are better at competing with only gain an increase in their own productivity, but other groups or deterring predators (Kokko et al. 2001 ; also gain the indirect benei ts associated with subordi- Clutton-Brock 2002 , 2009 ). Under this scenario, helpers nate females breeding (Richardson et al. 2002 ). In con- not only gain benei ts for themselves, but also benei t trast, unassisted dominant females on medium-quality other individuals in the breeding group. territories produced sex ratios around parity, and on In the Seychelles warbler, helpers have both a short- low-quality territories, where the presence of subordi- and long-term ef ect on of spring i tness: the number of nates is costly for breeding pairs because of competi- helpers in a territory is positively associated with both tion for food, unassisted dominant females maximize

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 208208 99/28/2015/28/2015 3:38:263:38:26 PPMM Seychelles warblers 209

(a) (b)

n = 276 73 11 n = 276 45 6 1.0

0.8

Annual survival 0.6

Adult 0.4 Juvenile

0 12 012 No. of helpers foster territory No. non-helping subordinates foster territory

Figure 12.10. Annual adult and juvenile survival probabilities (± S.E.) in relation to (a) the number of helpers in the rearing territory, and (b) the number of nonhelpers in the rearing territory. n denotes the number of individuals that were monitored for survival in their i rst year of life and subsequently for annual adult survival. From Brouwer et al. ( 2012).

1.0

0.9 1.0

0.8 0.8

Annual survival 0.7 0.6 0.6

1234560.4 Group size

Figure 12.11. A d u l t s u r v i v a l ( ± 9 5 % c o n i dence interval) prob- 0.2 1994 Proportion male nestlings abilities for an average year (thick lines) in relation to average 1993 group size inhabited as an adult. From Brouwer et al. ( 2006 ). 0.0 1995 Low Medium High their i tness by biasing the egg sex-ratio toward males Territory quality ( Figure 12.12 ), thus producing a greater proportion of Figure 12.12. Sex ratio of nestlings produced by pairs in male of spring that will disperse from the natal territory relation to quality of breeding territory (low-, medium- and and avoid competing for food with the breeding pair. high-quality territories; 1993–1995). Young were hatched from h is ability to bias the sex ratio, and its link to terri- one-egg clutches only. No additional young were present on tory quality, was coni rmed during the establishment the territory. From Komdeur et al. (1997 ).

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 209209 99/28/2015/28/2015 3:38:263:38:26 PPMM 210 Jan Komdeur et al.

of a new population on Aride Island. Breeding pairs n = 24 8 410 that were transferred from low- to high-quality terri- 0.8 p = 0.014 tories switched from producing 90% sons to produc- ing 85% daughters. Breeding pairs that were switched between high-quality territories showed no change in 0.6 sex ratio, producing 80% daughters before and after the switch (Komdeur et al. 1997 ). Sex-specii c embry- onic mortality between egg laying and hatching can 0.4 be ruled out, because the sexes of dead embryos were not biased toward the less adaptive sex (Komdeur et al. p 1997, 2002). 0.2 = 0.026 We have been able to experimentally coni rm the Benefits (offspring equivalents) dif ering i tness consequences of unassisted breed- ing pairs producing daughters or sons on high- and 0.0 Son Daughter Son Daughter low-quality territories, respectively. Young nestlings LOW HIGH were cross-fostered between unassisted breeding pairs on high- and low-quality territories that were feeding a Territory quality nestling of the putatively adaptive sex. Pairs breeding Figure 12.13. I n c l u s i v e i tness benei ts accrued to focal on high-quality territories that were allocated foster breeding pairs by producing sons or daughters (1985–1986) daughters gained signii cantly higher inclusive i tness on low- and high-quality territories. Fitness benei ts (± S.D.; benei ts than those raising foster sons, whereas the indirect = dark columns, direct = pale columns) were reverse was true for pairs breeding on low-quality ter- expressed in terms of additional number of yearlings produced ritories ( Figure 12.13 ). h ese i ndings provide strong by the focal breeding pair through help of their i rst-generation evidence that sex allocation in the Seychelles warbler is of spring. h e dispersal tactics and reproduction of all i rst-generation of spring were recorded throughout their lives. adaptive to the breeding pair. P v a l u e s f o r d i f erences between sons and daughters were determined by two-tailed Mann–Whitney U t e s t . n d e n o t e s number of breeding pairs. Analyses from Komdeur ( 1998 ). An alternative route to cooperative breeding deposed females left the territory to become l oaters, In cooperative breeding species helpers are often but 68% remained as subordinate females if, appar- younger, pre-reproductive individuals that have not ently, the quality of the territory was sui ciently high yet acquired an independent breeding position (Brown to support another subordinate. Of the demoted sub- 1 9 8 7; C o c k b u r n 1 9 9 8; K o e n i g a n d D i c k i n s o n 2 0 0 4; ordinates, 64% became helpers for the new dominant West et al. 2007 ). In a few long-lived mammalian spe- female. As of yet, no clear direct benei ts have been cies, however, older post-reproductive individuals identii ed as being gained by these deposed females. have been documented to provide kin-directed coop- It is clear, however, that by helping these new, related erative behavior (Packer et al. 1998 ; Pavelka et al. dominant females to reproduce, they can gain indirect 2002 ). In the Seychelles warbler, we found that some benei ts (Richardson et al. 2007 ). dominant breeders, mainly females, were deposed In the few other species (including humans) where from their breeding position (Richardson et al . 2007 ). post-reproductive helping occurs, the cessation of Importantly, the new dominant females that replaced reproduction by older females appears to be a strat- them were more closely related to the deposed females egy linked to senescence (Packer et al. 1998 ; Pavelka than one would expect by chance (mean r = 0.29); et al. 2002 ). h is does not appear to be the case in indeed, they are often their daughters. Some of these Seychelles warblers. Dominant breeders do not become

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 210210 99/28/2015/28/2015 3:38:273:38:27 PPMM Seychelles warblers 211

subordinates as part of a strategy linked to a senescent Understanding variation in natal dispersal reduction in their own reproductive output. Indeed between and within the sexes many demoted individuals are substantially younger Our study has demonstrated large spatial and tempo- (average age 5.2 years; Richardson et al. 2007 ) than the ral variation in the natal dispersal behavior of young age at which reproductive senescence occurs (from age Seychelles warblers. At the beginning of our study, 8 years onward, Hammers et al. 2012 ). young males dispersed from their natal territories more h e observation that adults that have lost a domi- often than females, but later in the study both males and nant breeding position help their younger relatives is females showed similar natal dispersal. h e observed important because this represents an alternative route change was the result of females starting to disperse to cooperative breeding. Becoming a helper is not necessarily an ef ective strategy for all adult females earlier and not the result of a change in the quality of challenged by a rival, however. h e ef ectiveness of the territories on which males and females were born. the strategy depends on the degree of relatedness We can think of several possible explanations for this between the demoted female and the new dominant change, but the most likely appears to be changes in i t- female. ness gained by females dispersing at younger ages. In A recent study suggests that an alternative respons e the beginning of the study period (1982–1990) none of might be available. If a female dominant allows a sub- the dispersing female subordinates were observed as ordinate related female helper to cobreed in the same l oaters and all settled as primary breeders (Komdeur nest she will gain indirect i tness benei ts comparable 1992 ), whereas later in the study (1995–2005) 16 % to a demoted nonreproducing female that becomes of dispersing female subordinates became l oaters a helper (Hammers et al. 2012 ), but in addition the (Eikenaar et al. 2007 ). h ese l oater females do not gain dominant female gains the direct i tness benei ts from parentage (there is no egg dumping), and are therefore laying her own egg. For a dominant female it would excluded from reproducing. h is increase in the num- then be better to allow another subordinate female ber of female l oaters over time suggests that females to become a joint-neste r than to be demoted and have more dii culty recruiting into the breeding pop- instead help a new unrelated or less-related dominant ulation and that the i tness gains from leaving the natal female. It remains to be investigated why this does territory may have declined over time. If this is true, not always happen and why some females become then the i tness gains from remaining on the natal ter- demoted helpers and not cobreed while others adopt ritory as a subordinate must also have declined, since the joint-nesting strategy. otherwise it is hard to explain the increased dispersal of females. To understand the evolution of delayed dispersal, it Conclusions and prospects is important to relate variation in timing of natal dis- persal to inclusive i tness benei ts gained over the life- Our study shows that the occurrence of group living time. h is would involve a comparison of i tness costs and helping behavior in Seychelles warblers can be and benei ts before dispersal (remaining on the natal explained by several ecological and social variables. territory as a subordinate) versus the costs and ben- We discussed the i tness implications resulting from ei ts after natal dispersal until death. Furthermore, these behaviors, not only for the subordinates but also experiments are needed to show a causal relationship for the dominants with which they share the territory. between the timing of natal dispersal and the resul- Moreover, we also demonstrated that the expression of tant i tness payof s. For example, natal dispersal could dispersal behavior and cooperative breeding change be delayed by decreasing the group size in which the over time, and dif erently for males and females. subordinate lives by removing individuals, or natal Despite the long-term nature of our study, however, we dispersal could be advanced by removing a parent are still left with many unresolved questions. (Eikenaar et al. 2007 ).

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 211211 99/28/2015/28/2015 3:38:273:38:27 PPMM 212 Jan Komdeur et al.

In cases where a parent was replaced by a stepparent territory at the earliest age possible, because every year that a either naturally or because of the removal of the parent male remains a subordinate is a year with no or very limited for translocation, young dispersed from the natal ter- reproduction. Given the importance of territory ownership, ritory earlier but were less likely to occupy a breeding competition over breeder vacancies may be more intense for males than for females. h is is supported by the fact that exper- vacancy in the next year (Eikenaar et al. 2007 ). h is may imentally created male vacancies were occupied faster than indicate that under such circumstances, subordinates female vacancies (Eikenaar et al. 2009 ). cannot postpone dispersal until the best moment, but may be forced out at suboptimal times. h e decision (2) Reduced costs of dispersing by males. h ere may be fewer to delay natal dispersal may therefore be an adaptive costs of dispersal for males than females, because males may be more readily accepted into new groups. However, it is strategy employed by the subordinate to wait until the unknown how or why group members may benei t from the chances of gaining a breeding position are high. presence of an immigrant male subordinate. Furthermore, when a parent was replaced by a step- parent, the likelihood of dispersal was higher for male (3) Higher i tness benei ts to males of joining other groups as than female subordinates (Eikenaar et al. 2007 ). h e subordinates. h e higher competition for male vacancies, and the limited i tness benei ts accrued from males remaining on logic behind this could be that the stepparent could the natal territory, might explain why a considerable fraction gain i tness benei ts if a female subordinate engages in of males disperse and become subordinates in other groups. joint nesting (Richardson et al. 2002 ). Stepparents may Males might disperse and join groups smaller than their natal evict unrelated subordinates, given the costs of group group to improve their survival prospects, or they might join living (Brouwer et al. 2006 ), but this counteracts the a group with an old dominant male to improve their chances observation that a dominant pair sometimes tolerates of reproducing with the dominant female while the dominant immigrant male subordinates. To investigate parental male is still alive, or in order to inherit the territory after his tolerance, detailed observations on agonistic inter- death. Males may move to other groups to avoid inbreeding actions between subordinates and (step)parents are in their natal territory, and choose to become subordinates required. in groups where they are less related to the females to avoid inbreeding when the males at a later stage become dominants in these group. However, there is as yet no evidence of inbreed- Understanding the evolution of variation ing avoidance (Richardson et al. 2004 ; Eikenaar et al. 2008 ). in helping behavior between and within Indeed, breeding between i rst-order relatives is relatively the sexes frequent, resulting in >5% of all of spring, despite negative i t- ness ef ects in poor-quality seasons (Richardson et al. 2004 ). Future studies should also focus on the evolutionary Clearly, we lack an understanding of what drives temporal var- causes of sex-biased helping. Our study has demon- iation in natal dispersal by males and females. strated that female Seychelles warblers gain more direct benei ts through cooperation than do males. h is may In the Seychelles warbler there is also variation explain the female bias in helping behavior observed among individuals in whether or not they provide help early in the study. What, however, are the factors that if there is an opportunity to do so, at least in females. drive males to disperse earlier and help less? In order Preliminary analyses suggest that subordinate females to answer this question we need to know what i tness that help produce more l edglings over their lifetime gains are made by early-dispersing males and females than subordinate females that do not help. h is was relative to the alternative of staying. not due to these helping female subordinates having We can suggest several ultimate explanations for the higher annual survival, but instead because they have observed sex dif erences: higher lifetime reproductive success than nonhelping (1) Territory ownership is more important for males than for female subordinates (H. Dugdale et al. unpubl. data). females. For males, this is a near prerequisite for reprodu c t i o n , Why, then, do not all subordinate females become whereas females can reproduce by joint nesting (Richardson helpers when an opportunity arises? h is may be et al. 2001, 2002). In other words, males should obtain a linked to their relatedness to the dominant female,

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 212212 99/28/2015/28/2015 3:38:283:38:28 PPMM Seychelles warblers 213

as discussed earlier, but there may also be other rea- Dowling, Jan Drachmann, Hannah Edwards, Cas sons. For example, it might be that only subordinates Eikenaar, Danni Gilroy, Frank Groenewoud, Martijn in sui ciently high condition can af ord to help. In Hammers, Ian Hartley, Kim Hutchins, Sjouke Kingma, the Seychelles warbler there is evidence that declines Kees van Oers, Lewis Spurgin, h or Veen, Jildou van in condition are associated with helping, and that the der Woude, David Wright, and many more. h e genetic expression of helping behavior is determined by body analyses undertaken at the University of Shei eld were condition (van de Crommenacker et al. 2011, 2012 ) . Up undertaken by Francine Jury and Patricia Salguiero, to now, condition-dependence in dispersal has rarely with the support of Deborah Dawson and Andy Krupa, been incorporated into models of helping strategies and at Groningen University were undertaken by Marco (Heinsohn and Legge 1999 ; Russell et al. 2003 ; Covas van der Velde. We also thank Sjouke Kingma, David and Griesser 2007 ; Lawson Handley and Perrin 2007 ). Wright, and Lewis Spurgin for their helpful comments Finally, in our studies of the Seychelles warbler we on this chapter. We are extremely grateful for the i nan- were able to estimate the direct lifetime i tness ben- cial support provided by the Netherlands Organization ei ts – the total number of l edglings produced – as a for Scientii c Research, the UK Natural Environmental result of helping. It is extremely dii cult, however, to Research Council, the Australian Research Council, assess all the inclusive i tness benei ts that are gained. the EU Marie Curie Fellowship Programme, the We have a good long-term database on individual-level Schure-Beijerinck-Popping Foundation, the Dobberke dispersal behavior and helping strategies, social and Foundation for Comparative Psychology, the Lucie environmental parameters, and a genetic pedigree. But Burgers Foundation for Comparative Behaviour it is still a very challenging and complex task to com- Research, and the Universities of East Anglia, pare an individual’s potential inclusive i tness benei t Groningen, and Shei eld. from dispersal versus that gained from remaining in the group and subsequently helping, or not. h is is because of hidden ef ects. For example, helping can also have REFERENCES long-term i tness ef ects for the of spring that have received helped (i.e., greater later-life survival; Brouwer Arnold , K. E. and Owens I. P. F. (1998 ). Cooperative breeding in et al. 2012 ), and living in groups itself can have costs birds: a comparative test of the life history hypothesis. Proc. R. Soc. London B , 265 , 739 – 745 . irrespective of helping. Nonetheless, the unique fea- Balshine-Earn , S. , Neat , F. C. , Reid, H. , and Taborsky , M. (1998 ). tures of the Seychelles warbler study of er the oppor- Paying to stay or paying to breed? Field evidence for direct tunity to answer this, and other, important questions. benei ts of helping behavior in a cooperatively breeding i sh. Behav. Ecol. , 9 , 432 – 458 . Barrett , E. L. B. , Burke , T. , Hammers, M. , Komdeur , J. , and Acknowledgments Richardson, D. S. (2013 ). Telomere length and dynamics predict mortality in a wild longitudinal study. Mol. Ecol., We are grateful to Nature Seychelles for allowing us 47, 41– 53. to work on Cousin Island, providing accommodation Bathe , G. M. and Bathe , H. V. (1982 ). Territory size and habitat and facilities and for collaborating with us on various requirements of the Seychelles Brush warbler Acrocephalus aspects of the warbler conservation and research. h e (Bebrornis) sechellensis. Cambridge: International Council for Bird Preservation, Internal Report. Department of Environment and the Seychelles Bureau Bennett , P. M. and Owens , I. P. F. (2002 ). Evolutionary Ecology of Standards gave permission for i eldwork and blood of Birds–Life Histories, Mating Systems, and Extinction. sampling. We thank numerous students and colleagues Oxford : Oxford University Press. for help in the i eld and for sharing their ideas, includ- Berg , E. C. , Eadie , J. M. , Langen, T. A., and Russell , A. F. ing Iain Barr, Emma Barrett, Dorte Bekkevold, Karen ( 2009 ). Reverse sex-biased philopatry in a cooperative Blaakmeer, Lyanne Brouwer, Tom Clarke, Corine Eising, bird: genetic consequences and a social cause . Mol. Ecol., Janske van de Crommenacker, Arjan Dekker, Damian 18 , 3486 – 3499 .

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 213213 99/28/2015/28/2015 3:38:283:38:28 PPMM 214 Jan Komdeur et al.

Brouwer , L. , Richardson, D. S. , Eikenaar , C. , and Komdeur , J. Eikenaar , C. , Komdeur , J. , and Rich ardson, D. S. (2008 ). (2006 ). h e role of group size and environmental factors on Natal dispersal patterns are not associated with in b r e e d - survival in a cooperatively breeding tropical passerine. J. ing avoidance in the Seychelles warbler. J. Evol. Biol., 21 , Anim. Ecol. , 75 , 1321 – 1329. 1106 – 1116 . Brouwer , L. , Tinbergen , J. M. , Both , C. , Bristol , R. , Eikenaar , C. , Richardson, D. S. , Brouwer , L. , Bristol , R. , and Richardson , D. S. , et al. ( 2009 ). Experimental evidence for Komdeur , J. (2009 ). Experimental evaluation of sex dif er- density-dependent reproduction in a cooperatively breed- ences in territory acquisition in a cooperatively breeding ing passerine . Ecology , 90 , 729 – 741 . bird . Behav. Ecol. , 20 , 207 – 214 . Brouwer , L. , Richardson , D. S. , and Komdeur J. ( 2012 ). Helpers Eikenaar , C. , Brouwer , L. , Komdeur, J. , and Richardson , D. S. at the nest improve late-life of spring performance: evi- ( 2010 ). Sex biased natal dispersal is not a i xed trait in a dence from a long-term study and a cross-foster experiment. stable population of Seychelles warblers . Behaviour , 147 , PLoS One , 7 , e33167 . 1577– 1590. Brown , J. L. ( 1978 ). Avian communal breeding systems . Annu. Ekman , J , Dickinson , J. L., Hatchwel l, B. J., and Griesser , Rev. Ecol. Syst. , 9 , 123 – 155 . M. (2004 ). Delayed dispersal . In: Ecology and Evolution Brown , J. L. (1987 ). Helping and Communal Breeding in of Cooperative Breeding in Birds, ed. W. D. Koenig and Birds: Ecology and Evolution. P r i n c e t o n , N J : P r i n c e t o n J. L. Dickinson . Cambridge: Cambridge University Press, University Press . pp. 35– 47. Clutton-Brock, T. H. (2002 ). Breeding together: kin selec- Emlen, S. T. (1982 ). h e evolution of helping. I. An ecological tion and mutualism in cooperative vertebrates . Science , constraints model . Am. Nat., 119 , 29 – 39. 296 , 69– 72. Emlen, S. T. (1991 ). h e evolution of cooperative breed- Clutton-Brock , T. ( 2009 ) Cooperation between non-kin in ani- ing in birds and mammals. In: Behavioural Ecology: an mal societies . Nature , 462 , 51– 57. Evolutionary Approach , 3rd edn, ed. J. R. Krebs and N. B. Cockburn , A. ( 1998 ). Evolution of helping behavior in coop- Davies . Oxford : Blackwell , pp. 301 – 337 . eratively breeding birds . Annu. Rev. Ecol. Syst. , 29 , 141 – 177 . Gaston , A. J. ( 1978 ). h e evolution of group territorial behavior Collar , N. J. and Stuart , S. N. ( 1985). h reatened Birds of Africa and cooperative breeding. Am. Nat., 112 , 1091 – 1100. and Related Islands . Cambridge : ICBP/IUCN Red Data Book Grii n , A. S. and West , S. A. (2002 ). : fact and i c- Part 1. International Council for Bird Preservation. tion . Trends. Ecol. Evol. , 17 , 15 – 21. Cornwallis , C. K. , West , S. A. , and Grii n , A. S. ( 2009 ). Routes Hadi eld, J. D. , Richardson, D. S. , and Burke , T. (2006 ). Towards to indirect i tness in cooperatively breeding vertebrates: kin unbiased parentage assignment: combining genetic, behav- discrimination and limited dispersal. J. Evol. Biol. , 22, ioural and spatial data in a Bayesian framework. Mol. Ecol., 2445– 2457. 15, 3715 – 3730. Covas , R. and Griesser , M. (2007 ). Life history and the evolu- Hammers , M. , Richardson , D. S., Burke , T. , and Komdeur , J. tion of family living in birds. Proc. R. Soc. London B , 274 , ( 2012 ). Age-dependent terminal declines in reproductive 1349– 1357. output in a wild bird. PLoS One, 7 , e40413 . Croft , D. P. , Krause , J. and James , R. ( 2004 ). Social networks in Hamilton, W. D. (1964 ). h e genetical evolution of social the guppy (Poecilia reticulata ). Biol. Lett ., 271 , 516 – 519 . behaviour I and II. J. h eor. Biol., 7 , 1–52 . Crook , J. ( 1960 ). h e Present Status of Certain Rare Land-birds of Hatchwell , B. J. and Komdeur, J. (200 0). Ecological constraints, the Seychelles Islands. Unnumbered Seychelles Government life history traits and the evolution of cooperative breeding. Bulletin, Department of the Environment, Victoria. Anim. Behav. , 59 , 1079 – 1086. Dickinson , J. L. and Hatchwell , B. J. (2004 ). h e i tness conse- Hatchwell , B. J. , Russell , A. F., Fow lie, M. K. , and Ross , D. J. quences of helping. In: Ecology and Evolution of Cooperative (1999 ). Reproductive success and nest-site selection in a Breeding in Birds , ed. W. D. Koenig and J. L. Dickinson . cooperative breeder: ef ect of experience and a direct ben- Cambridge : Cambridge University Press , pp. 48 – 66 . ei t of helping . Auk , 116 , 355 – 363 . Dickinson , J. L. , Koenig , W. D. , and Pitelka , F. A. ( 1996 ). Fitness Heinsohn , R. G. and Legge, S. (1999 ) . h e cost of helping . consequences of helping behavior in the western bluebird . Trends Ecol. Evol. , 14 , 53– 57. Behav. Ecol. , 7 , 168 – 177 . I U C N ( 2 0 1 3 ) . IUCN Red List of h reatened Species. Version Eikenaar , C. , Richardson , D. S. , Brouwer , L. , and Komdeur , J. 2013.2. www.iucnredlist.org . ( 2007 ). Parent presence, delayed dispersal, and terri- Koenig , W. D. and Dickinson, J. L., eds. (2004 ). Ecology tory acquisition in the Seychelles warbler. Behav. Ecol. , 18 , and Evolution of Cooperative Breeding in Birds. 874 – 879 . Cambridge : Cambridge University Press.

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 214214 99/28/2015/28/2015 3:38:283:38:28 PPMM Seychelles warblers 215

Koenig , W. D. , Pitelka , F. A. , Carmen, W. J. , Mumme , R. L. , and Komdeur , J. , Eikenaar , C. , Brouwer, L. , and Richardson , Stanback , M. T. ( 1992 ). h e evolution of delayed dispersal in D. S. (2008 ). h e evolution and ecology of cooperative cooperative breeders . Q. Rev. Biol. , 67 , 111 – 150 . breeding in vertebrates. In: Encyclopedia of Life Sciences. Kokko , H. , Johnstone , R. A. , and Clutton-Brock , T. H. ( 2001 ). Chichester : John Wiley and Sons, Ltd, pp. 1 – 8 . h e evolution of cooperative breeding through group aug- Lawson Handley, L. J. and Perrin, N. (2007 ). Advances in our mentation . Proc. R. Soc. London B , 268 , 187 – 196 . understanding of mammalian sex-biased dispersal . Mol. Komdeur , J. ( 1992 ). Importance of habitat saturation and ter- Ecol ., 16, 1559 – 1578. ritory quality for evolution of cooperative breeding in the Leadbeater , E. , Carruthers , J. M. , Green , J. P. , Rosser , N. S. , and Seychelles warbler . Nature , 358 , 493 – 495 . Field , J. ( 2011 ). Nest inheritance is the missing source of Komdeur , J. ( 1994a ). Conserving the Seychelles warbler direct i tness in a primitively eusocial insect. Science, 333 , Acrocephalus sechellensis by translocation from Cousin 874 – 876 . Island to the islands of Aride and Cousine. Biol. Conserv. , 76, Lessells , C. M. ( 1991 ). h e evolution of life histories . 143 – 152 . In: Behavioural Ecology: An Evolutionary Approach , 3 r d e d n , Komdeur , J. (1994b ). Experimental evidence for helping and ed. J. R. Krebs and N. B. Davies. Oxford : Blackwell , pp. 32 – 68 . hindering by previous of spring in the cooperative-breeding Loustau-Lalanne , P. ( 1968 ). h e Seychelles, Cousin Island Seychelles warbler Acrocephalus sechellensis . Behav. Ecol. Nature Reserve . Cambridge : International Council for Bird Sociobiol ., 34 , 175 – 186 . Preservation , Internal report. Komdeur , J. (1996a ). Seasonal timing of reproduction in a trop- Orchard , D. ( 2004 ). Aride Seychelles warbler Census. ical bird, the Seychelles warbler: a i eld experiment using In: Annual Report 2003. Aride Island Nature Reserve, translocation . J. Biol. Rhythms , 11 , 333 – 346 . Seychelles , ed. M. Betts. Royal Society of Wildlife Trusts, Komdeur , J. (1996b ). Facultative sex ratio bias in the of spring unpublished report. of Seychelles warblers. Proc. R. Soc. London B , 263 , 661 – 666 . Oustalet , M. E. (1878 ). Etude sur la faune omithologique des Komdeur , J. (1996c ). Inl uence of helping and breeding expe- isles Seychelles . Bull. Soc. Philomath. Paris, 7 , 161 – 206 . rience on reproductive performance in the Seychelles war- Packer , C. , Tatar , M. , and Collins, A. (1998 ). Reproductive ces- bler: a translocation experiment . Behav. Ecol. , 7 , 326 – 333 . sation in female mammals . Nature, 392 , 807 – 811 . Komdeur , J. ( 1998 ). Long-term i tness benei ts of egg sex modi- Pavelka , M. S. M. , Fedigan , L. M., and Zohar , S. (2002 ). i cation by the Seychelles warbler . Ecol. Lett. , 1 , 56– 62. Availability and adaptive value of reproductive and postre- Komdeur , J. ( 2003 ). Daughters on request: about helpers and egg productive Japanese macaque mothers and grandmothers. sexes in the Seychelles warbler. Proc. R. Soc. London B, 2 7 0, 3 – 1 1. Anim. Behav ., 64, 407 – 414 . Komdeur , J. and Edelaar , P. ( 2001 ). Male Seychelles warblers Richardson , D. S. , Jury , F. L. , Blaakmeer, K. , Komdeur , J. , and use territory budding to maximize lifetime i tness in a satu- Burke , T. ( 2001 ). Parentage assignment and extra-group rated environment . Behav. Ecol. , 12 , 706 – 715 . paternity in a cooperative breeder: the Seychelles warbler Komdeur , J. and Pels , M. D. ( 2005 ). Rescue of the Seychelles (Acrocephalus sechellensis ). Mol. Ecol., 10 , 2263 – 2273. warbler on Cousin Island, Seychelles: the role of habitat res- Richardson , D. S , Burke , T. , and Komde ur, J. (2002 ). Direct toration . Biol. Conserv. , 124 , 305 – 313 . benei ts explain the evolution of female biased cooper- Komdeur , J. , Huf stadt , A. , Prast , W. , Castle , G. , Mileto , R. , et al. ative breeding in the Seychelles warblers. Evolution , 56 , (1995 ). Transfer experiments of Seychelles warblers to new 2313– 2321. islands: changes in dispersal and helping behaviour. Anim. Richardson , D. S. , Komdeur , J. , and B urke, T. (2003a ). Altruism Behav. , 49 , 695 – 708 . and ini delity among warblers. Nature , 422 , 580. Komdeur , J. , Daan , S. , Tinbergen , J. , and Mateman , C. ( 1997 ). Richardson , D. S. , Burke , T. , and Kom deur, J. (2003b ). Extreme adaptive modii cation in sex ratio of the Seychelles S e x - s p e c i i c associative learning cues and inclusive i t- warbler’s eggs . Nature , 385 , 522 – 525 . ness benei ts in the Seychelles warbler. J. Evol. Biol., 16 , Komdeur , J. , Piersma , T. , Kraaijeveld , K. , Kraaijeveld-Smit , 854 – 861 . F. , and Richardson , D. S. ( 2004a ). Why Seychelles warblers Richardson , D. S. , Komdeur , J. , and B urke, T. (2004 ). Inbreeding fail to recolonize nearby islands: unwilling or unable to l y in the Seychelles warbler: environment-dependent mater- there? Ibis , 146 , 298 – 302 . nal ef ects . Evolution , 58 , 2037 – 2048. Komdeur , J. , Burke , T. , and Richardson , D. S. ( 2004b ). Richardson , D. S. , Bristol , R., and Shah, N. J. (2006 ). Experimental evidence that kin discrimination in the Translocation of the Seychelles warbler Acrocephalus Seychelles warbler is based on association and not on sechellensis to establish a new population on Denis Island, genetic relatedness . Proc. R. Soc. London B , 271 , 963 – 969 . Seychelles . Conserv. Evidence , 3 , 54 – 57.

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 215215 99/28/2015/28/2015 3:38:283:38:28 PPMM 216 Jan Komdeur et al.

Richardson , D. S. , Burke , T. , and Komdeur , J. ( 2007 ). Teunissen , N. ( 2013 ). A newly translocated population of the Grandparent helpers: the adaptive signii cance of older, Seychelles warbler: is there evidence for interspecii c com- postdominant helpers in the Seychelles warbler. Evolution , petition with Seychelles white-? M.Sc. thesis, Groningen 31, 2790 – 2800. University, h e Netherlands. Russell , A. F. , Sharpe , L. L. , Brotherton , P. N. M. , and Clutton-Brock , van de Crommenacker , J. and Richardson , D. S. (2007 ) T. H . ( 2 0 0 3 ) . C o s t m i n i m i z a t i o n b y h e l p e r s i n c o o p e rative verte- Monitoring and Studying the Seychelles warbler: i eldwork brates. Proc. Natl. Acad. Sci. (USA) , 100 , 3333 – 3338 . on Cousine Island. Report from the University of East Anglia, Saf ord , R. and Hawkins , F. ( 2013 ). h e Seychelles war- UK/ University of Groningen, Netherlands. bler . In: h e Birds of Africa: Volume VIII: h e Malagasy van de Crommenacker , J. , Komdeur, J. , and Richardson , D. S. Region: Madagascar, Seychelles, Comoros, Mascarenes . (2011 ). Assessing the cost of helping: the roles of body London : Christopher Helm , pp. 758 – 760 . condition and oxidative balance in the Seychelles warbler Spurgin , L. , Wright , D. J. , van der Velde , M. , Collar , N. J. , (Acrocephalus sechellensis) . PLoS One, 6 , e26423 . Komdeur , J. , et al. (2014 ). Museum DNA reveals the demo- van de Crommenacker , J. , Richardson , D. S., Koltz , A. M., graphic history of the endangered Seychelles warbler . Evol. Hutchings , K. , and Komdeur, J. (2012 ). Parasitic infection Applic. , 7 , 1134 – 1143. and oxidative stress are associated and vary with breeding Stacey , P. B. and Koenig , W. D. , eds. ( 1990 ). Cooperative activity in the Seychelles warbler. Proc. R. Soc. London B, Breeding in Birds: Long-term Studies of Ecology and Behavior. 279 , 1466 – 1476. Cambridge : Cambridge University Press . West , S. A. , Grii n A. S. , and Gardner , A. (2007 ). Evolutionary Stacey, P. B. and Ligon, J. D. (1987 ). Territory quality and dis- explanations for cooperation. Curr. Biol., 17 , 661 – 672 . persal options in the acorn woodpecker, and a challenge to Wright , D. J. , Shah , N. J. , and Richardson , D. S. (2014 ). the habitat-saturation model of cooperative breeding. Am. Translocation of the Seychelles warbler Acrocephalus Nat. , 130 , 654 – 676 . sechellensis to establish a new population on Frégate Island . Stacey , P. B. and Ligon , J. D. ( 1991) . h e benei ts-of-philopatry Conserv. Evidence , 11 , 20– 24. hypothesis for the evolution of cooperative breeding: varia- Zack , S. ( 1990 ). Coupling delayed breeding with short-distance tion in territory quality and group-size ef ects . Am. Nat , 137 , dispersal in cooperatively breeding birds. Ethology , 86 , 831 – 846 . 265 – 286 .

99781107043435c12_p197-216.indd781107043435c12_p197-216.indd 216216 99/28/2015/28/2015 3:38:293:38:29 PPMM