Journal of Animal Ecology 2014

doi: 10.1111/1365-2656.12206

Immigrants are attracted by local pre-breeders and recruits in a seabird colony K. Lesley Szostek1*, Michael Schaub2 and Peter H. Becker1 1 2

Institute of Avian Research “Vogelwarte Helgoland”, An der Vogelwarte 21, Wilhelmshaven, D-26386, Germany; and Swiss Ornithological Institute, Sempach, CH-6204, Switzerland

Summary 1. Immigration is a major demographic factor shaping population dynamics. However, due to methodological difficulties, the extent of immigration and factors affecting immigration are insufficiently studied. This is also true for seabird colonies. 2. We estimated annual immigration based on a long-term study of a colony of common terns Sterna hirundo marked with transponders, using a Bayesian integrated population model that links colony size and productivity with individual life histories. 3. Strong annual fluctuations in the number of immigrants were found. To identify whether colony-specific covariates influenced immigration, we related the number of immigrants to various proxy variables for breeding site quality, specifically colony size, productivity, number of local subadults and local recruits. Numbers of local recruits and local subadults showed strong positive correlations with number of immigrants. 4. We found that variation in immigration rate had strongly contributed to variation in colony growth rate, more so than variation in local recruitment or adult survival. 5. Collectively, results suggest that immigration strongly affects colony growth rate, that the driving force behind immigration is natal dispersal and that immigrants were attracted by local recruits. Key-words: Bayesian modelling, common tern, conspecific attraction, dispersal, local recruitment, mate availability, public information, Sterna hirundo

Introduction Population growth is determined by the interplay of the demographic rates survival, productivity, emigration and immigration (e.g. Begon, Mortimer & Thompson 1996). While there are many empirical studies demonstrating that variation in survival and productivity are important determinants for population growth, the impact of immigration and emigration is less well understood (Clobert & Lebreton 1991). Due to methodological difficulties in estimating emigration and immigration, these demographic processes are often assumed to cancel each other out (e.g. Reid et al. 2003) and are not included in population models. Recently developed models allow the estimation of immigration, and there is growing evidence that immigration is a crucial demographic process (Ward 2005; Mayer, Schiegg & Pasinelli 2009; Schaub et al. 2010, 2012; Schaub, Jakober & Stauber 2013). *Correspondence author. E-mail: lesley.szostek@ifv-vogelwarte. de

Dispersal processes are the consequence of two main individual decisions, namely the decision whether to leave from the current site and conditional on leaving, the decision where to settle next. The individual decision to stay or leave has been widely studied (e.g. Greenwood 1980; Johnson 1986; Joly & Miaud 1989; Lewis 1995; Hueter et al. 2004; Stenhouse & Robertson 2005; Coulson & Coulson 2008; Van der Waal, Mosser & Packer 2009), while fewer studies have investigated the settlement decision (e.g. Danchin, Boulinier & Massot 1998; Dittmann, Zinsmeister & Becker 2005; Clobert et al. 2009; Teichroeb, Wikberg & Sicotte 2011). The latter is more difficult to study as it typically requires the study of individuals from several connected populations. Various factors might cause individuals to move between birthplace and first breeding site (natal dispersal) or between breeding locations (breeding dispersal). These include breeding-/roosting-site availability (Lewis 1995), habitat quality (Spinks, Jarvis & Bennett 2000), food availability (Oro, Cam & Mart
ınez-Abra
ın 2004), conspecific attraction (Stamps 1988), mate availability (Steiner & Gaston 2005;

© 2014 The Authors. Journal of Animal Ecology © 2014 British Ecological Society

2 K. L. Szostek, M. Schaub & P. H. Becker Teichroeb et al. 2011), evading competition with settled (older) individuals (Long et al. 2008) or inbreeding avoidance (Long et al. 2008). All of them could impact the decision to leave and where to settle. The decision to leave induces emigration at the population level, while the settlement decision results in immigration, both creating exchange of individuals and gene flow among populations (e.g. Becker et al. 2008a, 2008b; Bicknell et al. 2012; Ludwig & Becker 2012). If it is possible to identify factors that influence immigration at the population level, it could allow insights into the individuals’ settlement decision (cf. Frederiksen & Bregnballe 2001). Several hypotheses about the settlement decision have been formulated and tested. The conspecific attraction hypothesis states that individuals are more likely to establish at places where many conspecifics are present (Stamps 1988; Reed & Dobson 1993). Well-established, large populations seem to attract more immigrants (Oro & Ruxton 2001; Serrano & Tella 2003; Cam et al. 2004). The availability of mates, especially in species with sex-biased dispersal, might also contribute to a positive relationship between number of immigrants and population size (Danchin & Wagner 1997; Steiner & Gaston 2005; Gauthier, Milot & Weimerskirch 2010). Especially in species with sex-biased dispersal, immigration of the less philopatric sex is necessary to find a mate (Dale 2001). According to the ‘public information theory’, individuals chose sites based on information on habitat quality gained by observing conspecifics at an unknown site (Valone 1989, 2007; Valone & Templeton 2002; Doligez et al. 2003; Danchin et al. 2004; Clobert et al. 2009). This theory fares well in model simulations (Boulinier & Danchin 1997; Doligez et al. 2003; Enfj€all & Leimar 2009). Furthermore, it has been found that prospecting occurs, when the most information about habitat quality is available (Boulinier et al. 1996), and that breeding success influences breeding site selection (Doligez, Danchin & Clobert 2002; Aparicio, Bonal & Mu~ noz 2007; Boulinier et al. 2008). However, not all studies found positive relationships between breeding success and dispersal probability (Oro & Ruxton 2001; Cam et al. 2004). Immigration into established breeding grounds has also been suggested to be related to environmental qualities such as food availability (Valone & Templeton 2002; Oro et al. 2004) in the prospective new habitat. While most studies report very high rates of philopatry in colonial seabirds (Coulson 2001), it has been suggested that the role and extent of dispersal are often underestimated (Coulson & Coulson 2008), supported by studies that found low philopatry in some colonial seabird species (Chastel, Weimerskirch & Jouventin 1993; Frederiksen & Petersen 2000). We estimated immigration and analysed factors affecting immigration in a common tern (Sterna hirundo) colony (Banter See) to assess the impact of immigration on colony growth and to study causes for settlement decisions. Long-term monitoring of an entire colony allowed us to differentiate between local recruits,

breeders, non-breeders and immigrants to study their influence on colony growth. The specific aims of this study were (i) to estimate immigration to the common tern colony at Banter See for each sex using Bayesian integrated population modelling, (ii) to estimate the relative importance of immigration to the colony growth rate and (iii) to link fluctuations in immigration to demographic traits, to analyse what factors influenced immigration into this colony. The factors considered were colony size (conspecific attraction), number of local subadults (future availability of mates, conspecific attraction), number of local recruits (availability of mates, conspecific attraction), productivity in the focal year (indicator of habitat quality and current breeding conditions) and in the previous year (public information). All of these factors are indirect measures of the quality of a colony site, potentially used as cues by immigrants for their settling decision.

Material and methods study site and data collection ‘Banter See’ colony is located in an abandoned harbour in Wilhelmshaven on the German North Sea coast (53 °30′40″N, 08 °06′20″E). The terns breed on six concrete islands, identical in size and substrate, which are fitted with walls to protect the colony from flooding and terrestrial predators. During the study period from 1994 to 2010, colony size fluctuated between 90 and 530 breeding pairs. Since 1980, birds have been ringed, and since 1992, all fledglings and 101 adults were injected subcutaneously with a passive transponder (Trovan ID 100, 11 9 2 mm) that identified the individual when read by an antenna (Becker & Wendeln 1997). No adults were caught and marked after 1995. Therefore, a proportion of unmarked individuals was present in the colony, including immigrants and a decreasing proportion of unmarked locals (between 75 and 55 % of breeders were unmarked). An electronic surveillance system of antennas on elevated platforms automatically and remotely recorded individual attendance throughout the breeding season (Becker, Wendeln & Gonz
alez-Solis 2001). Antennas were also placed around all clutches to identify breeders. Individuals that were only recorded on a platform and not on a clutch were assumed to be nonbreeders that year. These were mainly local subadults prospecting the colony from the age of 2 years (Dittmann et al. 2005). Reliability of the detection system was high (Becker et al. 2008b), and reencounter probability of breeders was close to 1, indicating that site fidelity was high and breeding skips rare (Szostek & Becker 2012). The breeding site was visited every 2–3 days to determine laying date and individual reproductive success (Wagener 1998). Marked adults were sexed using behavioural observations, and since 1998, chicks were sexed using standard PCR methods (Becker & Wink 2003). Colony size counts were derived from the total number of observed first clutches in the colony and not based on subsamples. From this information, we constructed three data sets. First, we assembled multistate encounter histories for each individual based on capture-redetection data, where the states were indicative of the age and breeding status. In total, the capture histories of 3355 individuals were included (1718 females, of which 1697

© 2014 The Authors. Journal of Animal Ecology © 2014 British Ecological Society, Journal of Animal Ecology

Immigration in a common tern colony 3 were marked as nestling, the remaining as adult; 1637 males, of which 1618 were marked as nestling, the remaining as adult). Secondly, we recorded the total number of fledglings produced per year by 2- to 5-year-old, by 6- to 10-year-old and by older than 10-year-old females and males, respectively. Only clutches from pairs where the age of at least one parent was known were included, as reproductive success is age-dependent (Ezard, Becker & Coulson 2006; Rebke et al. 2010) and terns tend to breed age assortatively (Ludwig & Becker 2008). We also recorded the number of females and males that raised these fledglings. In total, we included records of 2274 broods where at least the parental female was identified and 2344 broods where at least the parental male was identified. Finally, the annual colony size (number of breeding pairs) over the 17 study years was recorded.

integrated population model We used an integrated population model (Besbeas et al. 2002; Schaub & Abadi 2011) fitted in the Bayesian framework to analyse three data sets simultaneously. Advantages of this method are a proper error propagation, increased precision of parameter estimates and, most importantly, the ability to estimate demographic parameters for which no explicit data were collected (Abadi et al. 2010; K
ery & Schaub 2012). Here, we were particularly interested in estimating the number of immigrated individuals per year. An integral part of the integrated population model was a demographic model that links population sizes with demographic rates, like a matrix projection model (Caswell 2001). Based on knowledge about the common tern’s life history (Becker & Ludwigs 2004; Ezard et al. 2006; Szostek & Becker 2012), we constructed a pre-breeding census model with a total of 19 states. The states were defined based on a combination of the breeding status of an individual and its age. We defined five states for individuals that had never reproduced, four states of first-time breeders, nine states of experienced breeders and one state for immigrants. Appendix S1 provides the life cycle graph (Fig. A.1) and state definitions (Table A.1). The number of individuals in a state in year t is a function of the number of individuals in other states in year t 1 and the demographic parameters. Our model needed age-specific probabilities of survival (3 age classes: 0– 2 years, 3 year, older than 3 years) and of first-time breeding (4 age classes: first-time breeder at ages 2, 3, 4 or 5 years, cf. Ludwigs & Becker 2002), age-specific productivity (3 age classes. 2– 5 years old, 6–10 years old, at least 11 years old, cf. Ezard, Becker & Coulson 2006; Rebke et al. 2010) and breeding propensity (see Appendix Table A.1 for a complete list of parameters). The integrated population model also included demographic stochasticity using binomial and Poisson distributions to specify the relationship between the number of individuals in the 19 states in year t and in year t 1 and environmental stochasticity by considering random year effects for all demographic parameters. Appendix S1 provides a detailed description of the model. To estimate parameters (number of individuals in each of the 19 states in each year, demographic parameters in each year), the joint likelihood of the integrated population model was analysed (K
ery & Schaub 2012). The joint likelihood is composed of the likelihood of (i) a state-space model for the population count data, (ii) a multistate capture–recapture model for the individual encounter histories and (iii) Poisson regression models for the data on productivity. Due to its flexibility, we used the Bayesian

approach to analyse the integrated population model. This approach combines the joint likelihood with prior probability distributions to obtain posterior distributions of the target parameters based on Bayes’ theorem. We specified vague priors for the parameters to be estimated (see Appendix Table A.2 for their specification). We applied Markov Chain Monte Carlo (MCMC) methods to simulate observations from the posterior distributions with software WinBUGS (Lunn et al. 2000) that was run from R with package R2WinBUGS (Sturtz, Ligges & Gelman 2005). We specified a burn-in of 50 000 and simulated 150 000 samples that were thinned by a factor of 12 and ran 3 chains with different starting values. Inference was therefore obtained from 25 000 samples of the posterior distributions. Convergence of the Markov chain was evaluated with the Brooks–Rubin–Gelman diagnostics (Brooks & Gelman 1998) and was satisfactory for the ^ setting that we described (R\1
05). BUGS code of the model is available in Appendix A.2. The integrated population model requires an assumption about the age of immigrants. As common terns tend to exhibit low rates of breeding dispersal (Austin 1949; Becker et al. 2008b; Szostek & Becker 2012; but see Austin 1951; Morris, Pekarik & Moore 2012), immigrants are likely first-time breeders and therefore between 2 and 6 years old (Ludwigs & Becker 2002; mean age of first breeding 3
7  0
1 years, n = 14 years, Szostek & Becker 2012). Because many individuals start reproducing at 3 years, we assumed in our integrated population model that all immigrants were 3 years old. To test whether the assumption about the age of immigrants had a significant effect on parameter estimates, we also constructed a model where all immigrants were assumed to be 6 years old. Finally, we fitted a model that does not include any immigration to determine whether the inclusion of immigration is important at all. We fitted these three models to the female and male data independently. The count data were the same for both sexes, but the individual encounter histories and productivity data were different. In each case, we assumed that the other sex is an unlimited resource. We performed two separate analyses, because we wanted to assess whether there were sex differences in parameters (particularly in immigration).

demographic contributions to colony growth To assess how much the temporal variation of three demographic processes (survival of experienced breeders, local recruitment and immigration) had contributed to the variation in colony growth, we correlated colony growth rate with immigration rate, local recruitment rate and adult survival probability (cf. Robinson et al. 2004; Schaub, Jakober & Stauber 2013). These were all derived parameters (for their definition and calculation, see Appendix S1) and were calculated under consideration of the uncertainties of parameters estimates. Thus, we computed the correlation coefficients in each posterior sample to get their posterior distributions and also calculated the probability that they were positive [P(r > 0)]. These calculations are presented for females only.

factors influencing the number of immigrants We used correlation analyses to assess whether the number of immigrants (It) was related to: (i) colony size excluding immigrants (total number of individuals present at the colony minus

© 2014 The Authors. Journal of Animal Ecology © 2014 British Ecological Society, Journal of Animal Ecology

4 K. L. Szostek, M. Schaub & P. H. Becker 600

(a)

Breeding females

500

400

300

200

100

Observed Model estimates assuming no immigration Model estimates assuming immigration

0 300

Individuals

that year’s immigrants), (ii) number of local subadults/pre-breeders, (iii) number of local recruits, (iv) productivity (number of fledglings per breeding pairs) in the focal year, (v) productivity in the previous year. Components of productivity were positively autocorrelated in time (Durbin–Watson d < 1
5 for number of eggs, hatching success and d < 2 for fledging success), a prerequisite of the public information hypothesis (Valone & Templeton 2002). A positive correlation of the number of immigrants with these quantities would support the hypotheses that immigrants are inclined to settle in the colony because (i) they prefer to settle in larger colonies (conspecific attraction), (ii) there are many unpaired individuals as potential mates, (iii) there were many breeding partners available, (iv) productivity in this colony is high (young immigrants breed later than established breeders, so they can gauge breeding conditions of the focal year, before attempting reproduction themselves – public information), (v) productivity was good when they prospected the colony the year before (public information). The considered quantities are derived parameters (see Appendix A.1.4). We also tested whether immigration was related to breeding success in adjacent colonies (Appendix S2). The focal variables colony size, numbers of local recruits and subadults as well as productivity are likely correlated, and consequently, correlations between the number of immigrants and a focal variable could be spurious. To avoid this, we computed partial correlation coefficients between the number of immigrants and the focal variables, thus assessing the association between the number of immigrants and a focal variable after removing effects of the other focal variables. We calculated partial correlations for males and females together.

(b)

Immigrants Local recruits Local subadults

200

100

0 1994

1996

1998

2000

2002

2004

2006

2008

2010

Years

Results population size and demographic parameters The integrated population model including immigration approximated the observed population counts very precisely (Fig. 1a). The model without immigration, however, resulted in predictions of population size that were much smaller than the actual population counts (Fig. 1a). There were only negligible differences between results from the models assuming different ages of immigrants (Appendix A.3), and we therefore only reported results assuming all immigrants were 3 years old. The integrated population model produced estimates of many demographic parameters (Appendix A.3), most of which did not differ between sexes (P < 0
90). However, apparent survival of subadults was higher in males than in females (Φsub: male = 0
92 (SD: 0
03); female = 0
83 (SD: 0
03); P(♂>♀) = 0
99), and the probabilities of first breeding at ages 4 and 5 years were higher in males than in females [a4: male = 0
61 (0
06), female = 0
39 (0
11), P(♂>♀) = 0
96; a5: male = 0
58 (0
09), female = 0
33 (0
06), P(♂>♀) >0
99)]. Mean immigration rates were not different between sexes (males = 0
16, (0
12), females = 0
17 (0
13), P (♂>♀) = 0
57). The estimated immigration rate of 0
16 means that when colony size is 100 breeding pairs, about 16 individuals of each sex are expected to immigrate into the population. The number of immigrants peaked in the years 2003, 2004 and 2010 (Fig. 1b). The trend in

Fig. 1. Population growth of the common tern colony at Banter See between 1994 and 2010. (Panel a): number of breeding females observed in the population, estimated assuming immigration and assuming no immigration. (Panel b) estimated number of immigrants, local recruits and local subadults from a model assuming immigration (includes males and females). Note difference in scaling on y-axes.

immigrant numbers follows that of the number of local subadults with a 1 year lag (cf. Fig. 1b).

impact of immigration on colony growth Colony growth rate was significantly correlated with immigration rate [r = 0
931; P(r > 0) = 1], suggesting that the variation in immigration had a strong impact on colony dynamics. Colony growth rate was also significantly correlated with recruitment rate [r = 0
627; P(r > 0) = 1], but less strongly. Finally, colony growth was not significantly correlated with adult survival probabilities [r = 0
326; P(r > 0) = 0
900; see Fig. 2].

factors affecting immigration There was a significant partial correlation between the number of immigrants and number of local recruits [r = 0
504; P(r > 0) = 0
998; Fig. 3a]. The partial correlation between the number of immigrants and the number

© 2014 The Authors. Journal of Animal Ecology © 2014 British Ecological Society, Journal of Animal Ecology

Immigration in a common tern colony 5 (a) 1·5

Colony growth rate

1·4

r = 0·931 P (r > 0) = 1

1·3 1·2 1·1 1·0 0·9 0·8 –0·1

0·0

0·1

0·2

0·3

0·4

0·5

Immigration rate (b) 1·5

Colony growth rate

1·4

r = 0·627 P (r > 0) = 1

Discussion

1·3 1·2 1·1 1·0 0·9 0·8 0·0

0·1

number of immigrants in year t was neither correlated with productivity in year t [r = 0
100; P(r > 0) = 0
243; Fig. 3d] nor with productivity in year t 1 [r = 0
263; P (r > 0) = 0
055; Fig. 3e]. If the three peak immigration years, which might exert excess leverage on the correlation, were excluded, number of immigrants was still positively correlated with number of local recruits [r = 0
562; P(r > 0) = 0
987], but not positively correlated with any of the other variables (local subadults: r = 0
317; P(r > 0) = 0
844; local subadults previous year: r = 0
148; P(r > 0) = 0
663; colony size excluding immigrants: r = 0
416; P(r > 0) = 0
079; productivity previous year: r = 0
020; P(r > 0) = 0
474; there was, however, a negative correlation with productivity in the focal year: r = 0
652; P(r > 0) = 0
004).

0·2

0·3

0·4

0·5

0·6

0·7

0·8

Using an integrated population model, we estimated demographic rates of a common tern colony, including immigration. We found that immigration was substantial in both sexes and that the temporal variation in immigration was the most important demographic driver of variation in colony growth. Immigrants seemed to be attracted to the colony by the presence of potential breeding partners (local recruits). Previous productivity did not impact on immigration.

Local recruitment rate

impact of immigration on colony growth

(c) 1·5

Colony growth rate

1·4

r = 0·326 P (r > 0) = 0·900

1·3 1·2 1·1 1·0 0·9 0·8 0·70

0·75

0·80

0·85

0·90

0·95

1·00

Adult survival rate Fig. 2. Correlation of colony growth rate with rates of (a) immigration, (b) local recruitment and (c) adult survival. The horizontal and vertical lines show the limits of the 95% confidence intervals. Given are the estimated correlation coefficients and the probability that these are positive. The parameter estimates were taken from the model that considers females and assuming that the age of immigrants is 3 years.

of subadults (r > 0) = 0
904; colony size (r > 0) = 0
719; subadults of (r > 0) = 0
765]

was also quite strong [r = 0
260; P Fig. 3b], while the partial correlation with excluding immigrants [r = 0
089; P Fig. 3c] and with the number of returned the previous year [r = 0
164; P was weak and not significant. The

The match between observed colony size and model prediction was very close when immigration was included in the model. In contrast, if immigration was not considered, the model could not predict colony size correctly and the residual observation error increased considerably. Temporal variation in colony growth rate was most strongly connected with variation in immigration, and to a lesser extent with variation in local recruitment. Variation in adult survival was not an important driver of colony dynamics (Ezard, Becker and Coulson 2006), because it was relatively constant over time (Fig. 2, Szostek & Becker 2012). These results suggest that immigration is a crucial demographic process and therefore cannot be disregarded when studying population dynamics in open systems. Mean immigration rate was about 0
16, but it varied between 0
03 and 0
35 (Fig. 2a). On average, 70% (range: 38
2–88
9%) of recruits were immigrants. In some other studies on similar species, the percentage of immigrants among recruits was similar to that at our site, for example, at two colonies of black-legged kittiwakes Rissa tridactyla, an average of 77% and 96% of recruits were non-local (Coulson & Coulson 2008), while in a black guillemot Cepphus grylle colony between 53 and 63% of recruits was estimated to be immigrants (Frederiksen & Petersen 2000). The immigration rate we determined might well reflect the normal base-level dispersal between

© 2014 The Authors. Journal of Animal Ecology © 2014 British Ecological Society, Journal of Animal Ecology

6 K. L. Szostek, M. Schaub & P. H. Becker 350

r = 0·504 P (r > 0) = 0·998

300

Immigrants

250 200 150 100 50 0

(a)

–50 0

20

40

60

80

100

120

140

Local recruits 350

r = 0·089 P (r > 0) = 0·719

r = 0·269 P (r > 0) = 0·904

300

Immigrants

250 200 150 100 50

0

(b)

(c)

–50 0

50

100

150

200

250

300

350

200

600

800

r = –0·100 P (r > 0) = 0·243

300

1000

1200

Colony size (excluding Immigrants)

Local subadults 350

400

r = –0·263 P (r > 0) = 0·055

Immigrants

250 200 150 100 50

0 –50 0·0

(d) 0·4

0·8

1·2

1·6

Productivity (focal year)

2·0

(e) 0·0

0·4

0·8

1·2

1·6

2·0

Productivity (previous year)

colonies. However, distinct peaks with very high immigration indicated that in some years for some reason, immigrants were attracted to the study colony far more than in other years.

sex differences in survival and dispersal In birds, females often exhibit higher dispersal rates than males (Greenwood 1980; Becker et al. 2008b). However, we found no significant difference in the immigration rate between the sexes. Apparent survival probabilities in juveniles were also similar in males and females, suggesting that emigration in those age classes is equal between sexes. This is because after their first home migration, most juveniles pass through a prospecting phase, visiting multiple potential breeding sites (Dittmann, Zinsmeister and Becker 2005), while permanent emigration seemed mainly to happen in the subadult age class (Austin 1949; Boulinier et al. 1996; Tims et al. 2004; Dittmann, Zinsmeister and Becker 2005; Becker et al. 2008b), when apparent survival probabilities were significantly lower in females than males (Appendix Table A.3). Dispersal at young ages is consistent with other seabird species, for example, Arctic tern Sterna paradisaea (Møller,

2·4

Fig. 3. Correlation between the number of immigrants and (a) the number of local recruits, (b) present local subadults, (c) colony size (excluding immigrants), (d) productivity of the focal and (e) of the previous year. The horizontal and vertical lines show the limits of the 95% confidence intervals. Given are the estimated correlation coefficients and the probability that they are positive. The parameter estimates were taken from the model that considers females and assuming that the age of immigrants is 3 years.

Flensted-Jensen & Mardal 2006), black guillemot (Frederiksen & Petersen 2000), wandering albatross Diomedea exulans (Gauthier et al. 2010). We suggest that sex-specific difference in apparent survival in the third-year group is due to female-biased emigration, rather than an actual lower survival probability in females. At an adult age, there was no difference in apparent survival between sexes (Appendix Table A.3, Nisbet & Cam 2002), which is consistent with low breeding dispersal (e.g. Becker & Ludwigs 2004).

factors affecting immigration rate Most previous studies on this subject concentrated on the effects of productivity and often found positive relations between habitats with high breeding success in the previous year and subsequent high immigration or settlement rate (Danchin et al. 1998; Doligez et al. 2003; Aparicio et al. 2007). Despite strong fluctuations in productivity between the years, there was no relationship between productivity in the previous breeding season and immigration (Fig. 3e). There are several conceivable explanations for this: Perhaps, productivity of the previous year is not a reliable enough indicator, as it may have been more

© 2014 The Authors. Journal of Animal Ecology © 2014 British Ecological Society, Journal of Animal Ecology

Immigration in a common tern colony 7 affected by environmental conditions than breeding site quality. We also found no positive relationship between productivity in the focal year and immigration (Fig. 3d), indicating that current breeding conditions were an unlikely driver of immigration. Another explanation could be that more immigrants lead to increased foraging competition, which reduced productivity due to food depletion and prolonged foraging trips (Lewis et al. 2001; Tella et al. 2001; Forero et al. 2002; Gill, Hatch & Lanctot 2002; Davoren & Montevecchi 2003; Szostek et al. 2014). It is also possible that the driving force was the (possibly low) breeding success in the colonies that were left in favour of Banter See colony, as some species disperse after breeding failure (Burger 1982; Danchin, Boulinier and Massot 1998; Haas 1998; Schmidt 2004; Schaub & Von Hirschheydt 2009), others however, do not (e.g. Murphy 1996; Bried & Jouventin 1999; Gauthier, Milot & Weimerskirch 2010). The possibility that immigration peaks came about simply because more potential immigrants were in the system due to good breeding success in other colonies was considered in Appendix S2. It appeared that breeding success, and therefore cohort size, was not synchronous between adjacent colonies and high breeding success elsewhere was not consistently linked with immigration to the Banter See colony (Becker 1998). Sudden abandonment of colonies, as sometimes happens in the common tern (e.g. Austin 1951; Becker & Ludwigs 2004; Morris, Pekarik & Moore 2012), could also increase the number of potential immigrants in certain years, but to our knowledge, there was no such event in adjacent colonies in our study period (Appendix S2). This suggests that there are factors specific to Banter See colony that regulates how many of the potential immigrants in the system actually settle there in each year. We observed that experienced breeders in Wadden Sea colonies were often reluctant to move, even after several years of breeding failure (Becker 1998; also Tims et al. 2004). Consistently, productivity at Banter See colony seemed to have no effect on immigration, suggesting the public information hypothesis can be rejected for our study colony. The other major factor often suggested to influence immigration is conspecific attraction. This seemed to fit the situation at the study colony better, as we found that the number of immigrants tended to be higher in years of large colony size (excluding immigrants) and when subadult numbers were high. Presence of subadults might be attractive for immigrants, as they are unpaired and therefore potential mates (Ludwigs & Becker 2002; Becker et al. 2008a). This might explain why the strongest relationship we found was between the number of local recruits and immigrants. High numbers of local recruits indicate high mate availability for immigrants, but also good environmental conditions for first-time breeders. Recruitment might be influenced by environmental conditions (Crespin et al. 2006; common tern: Szostek & Becker 2012), and as the majority of immigrants are likely first-time breeders, they probably reacted to good

breeding conditions in the same way as local recruits, causing recruitment peaks. We had no way of attaining the number of unmarked prospectors at the study colony, but we think it is likely that large crowds of local subadults also attract immigrant prospectors, which might recruit at the colony the following year, creating immigration peaks. Immigrant numbers were also strongly correlated with returned subadults of the previous year (cf. Fig. 1b), but as these are in large parts the same individuals that become recruits the next year, effects of these two ‘attraction variables’ are difficult to separate. However, the underlying mechanism, young birds searching for a colony with potential mates, is the same. Years of high recruitment following years of high breeding success are not synchronous over the subregion around our study colony (Becker 1998); therefore, this is unlikely to be the cause for the correlation between number of immigrants and number of recruits. The effects of recruits attracting immigrants cannot be separated from the opposite effect of immigration augmenting recruitment. It therefore cannot be definitively said that high numbers of recruits induce high immigration, as it is possible that high recruitment rates were at least in part also brought about by immigrants pairing with local recruits. However, the timeline of high numbers of returning subadults preceding a year of high immigration (Fig. 1b) indicates that immigrants prospecting at the colony respond to the potential for a peak recruitment year and decide to settle there, taking advantage of a site with abundant breeding partners. This strong relationship suggests that a major factor for settlement decisions at Banter See colony is the availability of mates. In accordance with our results, immigration rate was also found to be correlated with local recruitment in a colony of southern fulmars Fulmarus glacialoides (Jenouvrier, Barbraud & Weimerskirch 2003). Steiner & Gaston (2005) found that dispersing thick-billed murres Uria aalge had a higher probability of attaining an experienced mate. Potential mates are more abundant in larger colonies (Steiner & Gaston 2005; Gauthier, Milot and Weimerskirch 2010), so the principles of conspecific attraction and the need to find mates might be linked. Prebreeders from other colonies, possibly along the migration route and in prospecting distance, may then use the increased probability of finding a mate, in years when many potential local recruits attend Banter See colony and thereby create immigration waves in years of high local recruitment. Given the slight difference in sex ratio (higher proportion of local male recruits: Becker et al. 2008b) and recruitment age (females usually breed 0
4 years earlier: Becker et al. 2008b; also Appendix Table A.3), as well as male-biased subadult survival (Appendix Table A.3, Becker et al. 2008b), local male recruits in the study colony must partly rely on immigrants for partners, especially in years, when the potential for local recruitment is high,

© 2014 The Authors. Journal of Animal Ecology © 2014 British Ecological Society, Journal of Animal Ecology

8 K. L. Szostek, M. Schaub & P. H. Becker otherwise a proportion of males will remain unpaired (Dale 2001). We found local male recruits were more often mated with immigrant partners than local female recruits, indicating that immigration is important for local males particularly to find mates. Becker et al. (2008b) estimated that 19% of female common tern recruits per year at Banter See might be immigrants. In conclusion, we estimated immigration rate and its yearly variation in a common tern colony. Our results highlight the importance of including immigration in population analysis, due to its impact on colony growth rate. Among the tested factors potentially attracting immigrants to the colony site, conspecific attraction was the main driver for immigration. Specifically, immigrants using the presence of subadults and recruits as a cue for colony attractiveness was a new finding. It remains to be tested, if the same mechanisms are also at play in other species.

Acknowledgements We thank the many field workers, who have helped compile the data for this study, for their untiring commitment over the years. Two anonymous reviewers provided very helpful comments to the manuscript. Mellumrat eV, Eric Stienen and Jan van der Winden provided data and information about other tern colonies. This project was funded by Deutsche Forschungsgemeinschaft DFG (BE 916). Permissions were granted by the regional authorities ‘Bezirksregierung Weser-Ems, Stadt Wilhelmshaven’ and ‘Nds. Landesamt f€ ur Verbraucherschutz und Lebensmittelsicherheit Oldenburg’.

References Abadi, F., Gimenez, O., Ullrich, B., Arlettaz, R. & Schaub, M. (2010) Estimation of immigration rate using integrated population models. Journal of Applied Ecology, 47, 393–400. Aparicio, J., Bonal, R. & Mu~ noz, A. (2007) Experimental test on public information use in the colonial Lesser Kestrel. Evolutionary Ecology, 21, 783–800. Austin, O.L. (1949) Site Tenacity, a Behaviour Trait of the Common Tern (Sterna hirundo Linn.). Bird-Banding, 20, 1–39. Austin, O.L. (1951) Group adherence in the Common Tern. Bird-Banding, 22, 1–15. Becker, P.H. (1998) Long-term trends of breeding success in Common Terns Sterna hirundo in the Wadden Sea. Vogelwelt, 119, 223–234. Becker, P.H. & Ludwigs, J.-D. (2004) Sterna hirundo Common Tern. BWP, 6, 91–137. Becker, P.H. & Wendeln, H. (1997) A New Application for Transponders in Population Ecology of the Common Tern. The Condor, 99, 534–538. Becker, P.H., Wendeln, H. & Gonz
alez-Solis, J. (2001) Population dynamics, recruitment, individual quality and reproductive strategies in common terns sterna hirundo marked with transponders. Ardea, 89, 241–252. Becker, P.H. & Wink, M. (2003) Influences of sex, sex composition of brood and hatching order on mass growth in common terns Sterna hirundo. Behavioral Ecology and Sociobiology, 54, 136–146. Becker, P.H., Dittmann, T., Ludwigs, J.-D., Limmer, B., Ludwig, S.C., Bauch, C. et al. (2008a) Timing of initial arrival at the breeding site predicts age at first reproduction in a long-lived migratory bird. PNAS, 105, 12349–12352. Becker, P.H., Ezard, T.H.G., Ludwigs, J.-D., Sauer-G€ urth, H. & Wink, M. (2008b) Population sex ratio shift from fledging to recruitment: consequences for demography in a philopatric seabird. Oikos, 117, 60–68. Begon, M., Mortimer, M. & Thompson, D.J. (1996) Population Ecology, 3rd edn. Blackwell Science Ltd, Oxford. Besbeas, P., Freeman, S.N., Morgan, B.J.T. & Catchpole, E.A. (2002) Integrating Mark–Recapture–Recovery and Census Data to Estimate

Animal Abundance and Demographic Parameters. Biometrics, 58, 540–547. Bicknell, A.W.J., Knight, M.E., Bilton, D., Reid, J.B., Burke, T. & Votier, S.C. (2012) Population genetic structure and long-distance dispersal among seabird populations: implications for colony persistence. Molecular Ecology, 21, 2863–2876. Boulinier, T. & Danchin, E. (1997) The use of conspecific reproductive success for breeding patch selection in terrestrial migratory species. Evolutionary Ecology, 11, 505–517. Boulinier, T., Danchin, E., Monnat, J.-Y., Doutrelant, C. & Cadiou, B. (1996) Timing of Prospecting and the Value of Information in a Colonial Breeding Bird. Journal of Avian Biology, 27, 252–256. Boulinier, T., McCoy, K.D., Yoccoz, N.G., Gasparini, J. & Tveraa, T. (2008) Public information affects breeding dispersal in a colonial bird: kittiwakes cue on neighbours. Biology Letters, 4, 538–540. Bried, J. & Jouventin, P. (1999) Influence of breeding success on fidelity in long-lived birds: an experimental study. Journal of Avian Biology, 30, 392–398. Brooks, S.P. & Gelman, A. (1998) Alternative methods for monitoring convergence of iterative simulations. Journal of Computational and Graphical Statistics, 7, 434–455. Burger, J. (1982) The role of reproductive success in colony-site selection and abandonment in Black Skimmers (Rynchops niger). The Auk, 99, 109–115. Cam, E., Oro, D., Pradel, R. & Jimenez, J. (2004) Assessment of hypotheses about dispersal in a long-lived seabird using multistate capture– recapture models. Journal of Animal Ecology, 73, 723–736. Caswell, H. (2001) Matrix Population Models. Construction, Analysis and Interpretation. Sinauer Associates, Sunderland, MA. Chastel, O., Weimerskirch, H. & Jouventin, P. (1993) High annual variability in reproductive success and survival of an Antarctic seabird, the snow petrel Pagodroma nivea. Oecologia, 94, 278–285. Clobert, J. & Lebreton, J.D. (1991) Estimation of demographic parameters in bird populations. Bird Population Studies: Relevance to Conservation and Management (eds C.M. Perrins, J.D. Lebreton & G.J.M. Hirons), pp. 75–104, Oxford University Press, New York. Clobert, J., Le Galliard, J.-F., Cote, J., Meylan, S. & Massot, M. (2009) Informed dispersal, heterogeneity in animal dispersal syndromes and the dynamics of spatially structured populations. Ecology Letters, 12, 197–209. Coulson, J.C. (2001) Colonial breeding in seabirds. Biology of Marine Birds (eds E.A. Schreiber & J. Burger), pp. 87–113. CRC Press, Boca Raton. Coulson, J.C. & Coulson, B.A. (2008) Measuring immigration and philopatry in seabirds; recruitment to Black-legged Kittiwake colonies. Ibis, 150, 288–299. Crespin, L., Harris, M.P., Lebreton, J.-D., Frederiksen, M. & Wanless, S. (2006) Recruitment to a seabird population depends on environmental factors and on population size. Journal of Animal Ecology, 75, 228–238. Dale, S. (2001) Female-biased dispersal, low female recruitment, unpaired males, and the extinction of small and isolated bird populations. Oikos, 92, 344–356. Danchin, E., Boulinier, T. & Massot, M. (1998) Conspecific reproductive success and breeding habitat selection: implication for the study of coloniality. Ecology, 79, 2415–2428. Danchin, E. & Wagner, R.H. (1997) The evolution of coloniality: the emergence of new perspectives. Trends in Ecology & Evolution, 12, 342–347.
Giraldeau, L.-A., Valone, T.J. & Wagner, R.H. (2004) Public Danchin, E., Information: from Nosy Neighbors to Cultural Evolution. Science, 305, 487–491. Davoren, G.K. & Montevecchi, W.A. (2003) Consequences of foraging trip duration on provisioning behaviour and fledging condition of common murres Uria aalge. Journal of Avian Biology, 34, 44–53. Dittmann, T., Zinsmeister, D. & Becker, P.H. (2005) Dispersal decisions: common terns, Sterna hirundo, choose between colonies during prospecting. Animal Behaviour, 70, 13–20. Doligez, B., Danchin, E. & Clobert, J. (2002) Public Information and Breeding Habitat Selection in a Wild Bird Population. Science, 297, 1107–1108. Doligez, B., Cadet, C., Danchin, E. & Boulinier, T. (2003) When to use public information for breeding habitat selection? The role of environmental predictability and density dependence. Animal Behaviour, 66, 973–988.

© 2014 The Authors. Journal of Animal Ecology © 2014 British Ecological Society, Journal of Animal Ecology

Immigration in a common tern colony 9 Enfj€ all, K. & Leimar, O. (2009) The evolution of dispersal – the importance of information about population density and habitat characteristics. Oikos, 118, 291–299. Ezard, T.H.G., Becker, P.H. & Coulson, T. (2006) The contributions of age and sex to variation in common tern population growth rate. Journal of Animal Ecology, 75, 1379–1386. Forero, M.G., Tella, J.L., Hobson, K.A., Bertellotti, M. & Blanco, G. (2002) Conspecific food competition explains variability in colony size: a test in Magellanic Penguins. Ecology, 83, 3466–3475. Frederiksen, M. & Bregnballe, T. (2001) Conspecific reproductive success affects age of recruitment in a great cormorant, Phalacrocorax carbo sinensis, colony. Proceedings of the Royal Society B, 268, 1519–1526. Frederiksen, M. & Petersen, E. (2000) The importance of natal dispersal in a colonial seabird, the Black Guillemot Cepphus grylle. Ibis, 142, 48–57. Gauthier, G., Milot, E. & Weimerskirch, H. (2010) Small-scale dispersal and survival in a long-lived seabird, the wandering albatross. Journal of Animal Ecology, 79, 879–887. Gill, V.A., Hatch, S.A. & Lanctot, R.B. (2002) Sensitivity of breeding parameters to food supply in Black-legged Kittiwakes Rissa tridactyla. Ibis, 144, 268–283. Greenwood, P.J. (1980) Mating systems, philopatry and dispersal in birds and mammals. Animal Behaviour, 28, 1140–1162. Haas, C. (1998) Effects of Prior Nesting Success on Site Fidelity and Breeding Dispersal: an Experimental Approach. The Auk, 115, 929–936. Hueter, R.E., Heupel, M.R., Heist, E.J. & Keeney, D.B. (2004) Evidence of philopatry in sharks and implications for the management of shark fisheries. Journal of Northwest Atlantic Fishery Science, 35, 239–247. Jenouvrier, S., Barbraud, C. & Weimerskirch, H. (2003) Effects of climate variability on the temporal population dynamics of southern fulmars. Journal of Animal Ecology, 72, 576–587. Johnson, C.N. (1986) Sex-biased philopatry and dispersal in mammals. Oecologia, 69, 626–627. Joly, P. & Miaud, C. (1989) Fidelity to the breeding site in the alpine newt Triturus alpestris. Behavioural Processes, 19, 47–56. K
ery, M. & Schaub, M. (2012) Bayesian Population Analysis Using WinBUGS. A Hierarchical Perspective. 1st edn. Academic Press, Waltham. Lewis, S.E. (1995) Roost fidelity of bats: a review. American Society of Mammalogists, 76, 481–496. Lewis, S., Sherratt, T.N., Hamer, K.C. & Wanless, S. (2001) Evidence of intra-specific competition for food in a pelagic seabird. Nature, 412, 816–819. Long, E.S., Diefenbach, D.R., Rosenberry, C.S. & Wallingford, B.D. (2008) Multiple proximate and ultimate causes of natal dispersal in white-tailed deer. Behavioral Ecology, 19, 1235–1242. Ludwig, S. & Becker, P. (2008) Supply and demand: causes and consequences of assortative mating in common terns Sterna hirundo. Behavioral Ecology and Sociobiology, 62, 1601–1611. Ludwig, S.C. & Becker, P.H. (2012) Immigration prevents inbreeding in a growing colony of a long-lived and philopatric seabird. Ibis, 154, 74–84. Ludwigs, J.-D. & Becker, P.H. (2002) The hurdle of recruitment: influences of arrival date, colony experience and sex in the common tern Sterna hirundo. Ardea, 90, 389–399. Lunn, D.J., Thomas, A., Best, N. & Spiegelhalter, D. (2000) WinBUGS – A Bayesian modelling framework: concepts, structure and extensibility. Statistics and Computing, 10, 325–337. Mayer, C., Schiegg, K. & Pasinelli, G. (2009) Patchy population structure in a short-distance migrant: evidence from genetic and demographic data. Molecular Ecology, 18, 2353–2364. Møller, A.P., Flensted-Jensen, E. & Mardal, W. (2006) Dispersal and climate change: a case study of the Arctic tern Sterna paradisaea. Global Change Biology, 12, 2005–2013. Morris, R.D., Pekarik, C. & Moore, D.J. (2012) Current status and abundance trends of Common Terns breeding at known coastal and inland nesting regions in Canada. Waterbirds, 35, 194–207. Murphy, M.T. (1996) Survivorship, breeding dispersal and mate fidelity in Eastern Kingbirds. The Condor, 98, 82–92. Nisbet, I.C.T. & Cam, E. (2002) Test for age-specificity in survival of the common tern. Journal of Applied Statistics, 29, 65–83. Oro, D., Cam, E. & Mart
ınez-Abra
ın, A. (2004) Influence of food availability on demography and local population dynamics in a long-lived seabird. Proceedings of the Royal Society B, 271, 387–396. Oro, D. & Ruxton, G.D. (2001) The formation and growth of seabird colonies: Audouin’s gull as a case study. Journal of Animal Ecology, 70, 527–535.

Rebke, M., Coulson, T., Becker, P.H. & Vaupel, J.W. (2010) Reproductive improvement and senescence in a long-lived bird. PNAS, 107, 7841–7846. Reed, J.M. & Dobson, A.P. (1993) Behavioural constraints and conservation biology: conspecific attraction and recruitment. Trends in Ecology & Evolution, 8, 253–256. Reid, J.M., Bignal, E.M., Bignal, S., McCracken, D.I. & Monaghan, P. (2003) Environmental variability, life-history covariation and cohort effects in the red-billed chough Pyrrhocorax pyrrhocorax. Journal of Animal Ecology, 72, 36–46. Robinson, R.A., Green, R.E., Baillie, S.R., Peach, W.J. & Thomson, D.L. (2004) Demographic mechanisms of the population decline of the song thrush Turdus philomelos in Britain. Journal of Animal Ecology, 73, 670–682. Schaub, M. & Abadi, F. (2011) Integrated population models: a novel analysis framework for deeper insights into population dynamics. Journal of Ornithology, 152, 1–11. Schaub, M., Jakober, H. & Stauber, W. (2013) Strong contribution of immigration to local population regulation : evidence from a migratory passerine. Ecology, 94, 1828–1838. Schaub, M. & Von Hirschheydt, J. (2009) Effect of current reproduction on apparent survival, breeding dispersal, and future reproduction in barn swallows assessed by multistate capture–recapture models. Journal of Animal Ecology, 78, 625–635. Schaub, M., Aebischer, A., Gimenez, O., Berger, S. & Arlettaz, R. (2010) Massive immigration balances high anthropogenic mortality in a stable eagle owl population: lessons for conservation. Biological Conservation, 143, 1911–1918. Schaub, M., Reichlin, T., Abadi, F., K
ery, M., Jenni, L. & Arlettaz, R. (2012) The demographic drivers of local population dynamics in two rare migratory birds. Oecologia, 168, 97–108. Schmidt, K.A. (2004) Site fidelity in temporally correlated environments enhances population persistence. Ecology Letters, 7, 176–184. Serrano, D. & Tella, J.L. (2003) Dispersal within a spatially structured population of lesser kestrels: the role of spatial isolation and conspecific attraction. Journal of Animal Ecology, 72, 400–410. Spinks, A.C., Jarvis, J.U.M. & Bennett, N.C. (2000) Comparative patterns of philopatry and dispersal in two common mole-rat populations: implications for the evolution of mole-rat sociality. Journal of Animal Ecology, 69, 224–234. Stamps, J.A. (1988) Conspecific attraction and aggregation in territorial species. The American Naturalist, 131, 329–347. Steiner, U.K. & Gaston, A.J. (2005) Reproductive consequences of natal dispersal in a highly philopatric seabird. Behavioral Ecology, 16, 634–639. Stenhouse, G.J. & Robertson, I.J. (2005) Philopatry, site tenacity, mate fidelity, and adult survival in sabine’s gulls. The Condor, 107, 416–423. Sturtz, S., Ligges, U. & Gelman, A. (2005) R2WinBUGS: a package for running WinBUGS from R. Journal for Statistical Software, 12, 1–16. Szostek, K.L. & Becker, P.H. (2012) Terns in trouble: demographic consequences of low breeding success and recruitment on a common tern population in the German Wadden Sea. Journal of Ornithology, 153, 313–326. Szostek, K.L., Becker, P.H., Meyer, B.C., Sudmann, S.R. & Zintl, H. (2014) Colony size and not nest density drives reproductive output in the Common Tern Sterna hirundo (ed R Robinson). Ibis, 156, 48–59. Teichroeb, J.A., Wikberg, E.C. & Sicotte, P. (2011) Dispersal in male ursine colobus monkeys (Colobus vellerosus): influence of age, rank and contact with other groups on dispersal decisions. Behaviour, 148, 765–793. Tella, J.L., Forero, M.G., Bertellotti, M., Don
azar, J.A., Blanco, G. & Ceballos, O. (2001) Offspring body condition and immunocompetence are negatively affected by high breeding densities in a colonial seabird: a multiscale approach. Proceedings of the Royal Society B, 268, 1455–1461. Tims, J., Nisbet, I.C.T., Friar, M.S., Mostello, C. & Hatch, J.J. (2004) Characteristics and Performance of Common Terns in Old and Newly-established Colonies. Waterbirds, 27, 321–332. Valone, T.J. (1989) Group foraging, public information, and patch estimation. Oikos, 56, 357–363. Valone, T. (2007) From eavesdropping on performance to copying the behavior of others: a review of public information use. Behavioral Ecology and Sociobiology, 62, 1–14.

© 2014 The Authors. Journal of Animal Ecology © 2014 British Ecological Society, Journal of Animal Ecology

10 K. L. Szostek, M. Schaub & P. H. Becker Valone, T.J. & Templeton, J.J. (2002) Public information for the assessment of quality: a widespread social phenomenon. Proceedings of the Royal Society B, 357, 1549–1557. Van der Waal, K.L., Mosser, A. & Packer, C. (2009) Optimal group size, dispersal decisions and postdispersal relationships in female African lions. Animal Behaviour, 77, 949–954. Wagener, M. (1998) Praktische Hinweise f€ ur brutbiologische Untersuchungen an der Flußseeschwalbe Sterna hirundo. Vogelwelt, 119, 279–286. Ward, M.P. (2005) The Role of Immigration in the Decline of an Isolated Migratory Bird Population. Conservation Biology, 19, 1528–1536. Received 21 May 2013; accepted 20 January 2014 Handling Editor: Henri Weimerskirch

Supporting Information Additional Supporting Information may be found in the online version of this article. Appendix S1. Detailed description of the integrated population model. Appendix S2. Comparative analysis of breeding success and colony size in common tern colonies in surrounding areas.

© 2014 The Authors. Journal of Animal Ecology © 2014 British Ecological Society, Journal of Animal Ecology