Skip to main content
NCBI home page
Proceedings of the Royal Society B: Biological Sciences logoLink to Proceedings of the Royal Society B: Biological Sciences
. 2008 Jun 3;275(1647):2063–2069. doi: 10.1098/rspb.2008.0356

Kin in space: social viscosity in a spatially and genetically substructured network

Jochen BW Wolf 1,2,*, Fritz Trillmich 2
PMCID: PMC2603206  PMID: 18522913

Abstract

Population substructuring is a fundamental aspect of animal societies. A growing number of theoretical studies recognize that who-meets-whom is not random, but rather determined by spatial relationships or illustrated by social networks. Structural properties of large highly dynamic social systems are notoriously difficult to unravel. Network approaches provide powerful ways to analyse the intricate relationships between social behaviour, dispersal strategies and genetic structure. Applying network analytical tools to a colony of the highly gregarious Galápagos sea lion (Zalophus wollebaeki), we find several genetic clusters that correspond to spatially determined ‘network communities’. Overall relatedness was low, and genetic structure in the network can be interpreted as an emergent property of philopatry and seems not to be primarily driven by targeted interactions among highly related individuals in family groups. Nevertheless, social relationships between directly adjacent individuals in the network were stronger among genetically more similar individuals. Taken together, these results suggest that even small differences in the degree of relatedness can influence behavioural decisions. This raises the fascinating prospect that kin selection may also apply to low levels of relatedness within densely packed animal groups where less obvious co-operative interactions such as increased tolerance and stress reduction are important.

Keywords: altruism, co-operation, cross-fostering, Zalophus wollebaeki, kin selection, social network

1. Introduction

Population substructuring is a key component in animal social systems. A growing number of theoretical studies recognize that who-meets-whom is not random, but determined by spatial relationships or other forms of assortment (Lion & van Baalen 2008). Social interactions on local and global population levels delimit the boundaries in which fitness-relevant behaviours are displayed, thus controlling patterns of gene flow (Sugg et al. 1996). Owing to philopatry and strategies of delayed dispersal (social viscosity), individuals are often physically assorted into kin clusters where animals share the same alleles by descent with an increased probability (Dobson 1982; Fowler 2005). For communally breeding species, where the average relatedness within these clusters is high and close relatives recognize each other, genetic structure is often dictated by the placement of related individuals and usually co-varies with social structure (Faulkes et al. 1997; Dobson 1998). For such systems, kin selection theory provides an appealing solution to understand how even costly social behaviours such as cooperation and altruism evolve as a direct evolutionary consequence of inclusive fitness benefits (Hamilton 1964; Griffin & West 2003). In large, colonially organized animal groups much less is known about the potential role of kin and the interplay between dispersal strategies, genetic structure and social behaviour. In such groupings, interactions between kin are diluted by the presence of numerous neighbours and average relatedness quickly drops with increasing group size (Lukas et al. 2005). Particularly in fission–fusion groups, where structural features are not intuitive, it is a major challenge to understand the assortative forces behind the structural properties of the system.

A promising approach is offered by the application of recent network theory to animal societies (e.g. Lusseau et al. 2005; Croft et al. 2006). This approach has the potential to explore apparently complex societies with seemingly ambiguous structure (Krause et al. 2007; Wey et al. 2008). Rooted in mathematical graph theory, network analysis can model the evolution of social behaviour (e.g. Ohtsuki et al. 2006; Taylor et al. 2007) and illustrate social tendencies with spatially explicit graphs. Parallel to the idea of social viscosity in kinship theory, most of the models expose population substructuring as the key element promoting the evolution of social behaviour (Lion & van Baalen 2008). Graph theoretical approaches and the traditional framework of inclusive fitness obey the same mathematical formalism and have other central elements in common (Lehmann et al. 2007; Taylor et al. 2007). Network approaches are thus appropriate for studying the relative role of kin selection in structured animal populations. While the theoretical literature on the subject is thriving, empirical studies only gradually appear to recognize the potential of social networks in addressing questions related to the evolution of social behaviour.

We have recently studied the structural properties of a highly dynamic fission–fusion social network that integrates proximity relationships of several hundred individuals (Wolf et al. 2007a). The network described in this study refers to a colony of sea lions (Zalophus wollebaeki; Wolf et al. 2007b) that dwell on a tiny islet (approx. 250 m wide) in the Galápagos archipelago. Approximately, 1000 individuals maintain social interactions year round (Wolf & Trillmich 2007). Newly developed algorithms screening the network's topology unveiled hierarchical elements of social structure consisting of intermediate tiers that would have otherwise gone undetected. Figure 1 shows a schematic of the network's essential structural properties: the overall social network is partitioned into large ‘communities’ and smaller ‘cliques’ nested therein. Community membership describes an individual's general social neighbourhood and can be explained by home range overlap (i.e. static interaction). Clique membership, however, represents an individual's closer social neighbourhood and cannot be explained spatially (i.e. dynamic interaction). The immediate social relationships of an individual are represented by the direct connections (ties) to other individuals in the network.

Figure 1.

Figure 1

Open in a new tab

Hierarchical structure in a social network of the Galápagos sea lion (modified after Wolf et al. 2007a). (a) Community structure in the overall social network. Nodes symbolize individual animals; lines stand for social ties. In the network, social entities can be identified, where members are more densely connected among themselves than with the rest of the network. They are called ‘communities’ and indicated by different colours. It is important to note that network structure is solely derived from social interaction data. (b) A map of the study island (approx. 250 m in width) shows that communities can largely be defined by patterns of space use. The coloured patches correspond to the home ranges of individuals from a given community showing that community members have common areas to which they are all faithful. (c) Communities further split into social cliques indicated by different symbols. Cliques represent an individual's most intimate social neighbourhood. In contrast to communities, the existence of cliques cannot be explained by patterns of space use. Home ranges of all clique members fully overlap regardless to which clique they belong.

As a high degree of fine-scale site fidelity and possibly female biased home range inheritance have been documented in this species (Wolf & Trillmich 2007) exploring this network holds the potential to work out the relationship between spatial, social and genetic structure. Specifically, genetic structure within a social group can emerge in two non-mutually exclusive ways which natural selection could act to favour: either as a by-product of philopatric behaviour per se or by selection on targeted kin interactions among small groups of highly related animals. The network approach allows us to make predictions on the relative role of the two processes.

  1. If genetic relatedness emerges primarily as a by-product of cross-generational site fidelity, then average relatedness among individuals across the colony as a whole should be generally low and genetic structure should be governed by the spatially defined communities within the network. Owing to philopatry, individuals of the same community would be expected to be more related to each other than to the global population background. Relatedness within cliques (the socially defined tier) should not be higher than in communities.

  2. However, if social ties are strongest among close kin that benefit from shared alleles via inclusive fitness, then there will be a higher degree of relatedness within the cliques than within the spatially defined communities. We may then expect to see cooperative or altruistic behaviour within cliques. One of the most conspicuous altruistic behaviours is cross fostering, which is widespread in some pinniped species despite its costs (e.g. Riedman & Le Boeuf 1982; Boness et al. 1998; Childerhouse & Gales 2001). Thus, we expect such behaviour to occur preferentially among highly related individuals from the same clique in the network.

2. Material and methods

This study is part of a long-term project conducted on a small Galápagos islet (Ecuador), which is used as a breeding and resting site by approximately 1000 sea lions (for details of the study area, study population, capture and tagging procedure, see Wolf et al. 2005 and Wolf & Trillmich 2007).

(a) Genetic and network analyses

The network part of this work is based on a study by Wolf et al. (2007a) on sea lion social behaviour witnessed during 10 September 2004–20 November 2004. For all 380 individuals forming part of the network, genomic DNA was amplified in a multiplex PCR approach (Qiagen multiplex PCR Kit). Samples were genotyped on an ABI 3730 capillary sequencer at 22 fluorescently marked dinucleotide microsatellite loci, most of which have been specifically designed for the Galápagos sea lion (Wolf et al. 2006; Hoffman et al. 2007). Individuals were only included in subsequent analyses if a minimum of 20 loci were successfully scored (Wolf et al. 2007b, 2008). Maximum-likelihood estimates of pairwise relatedness coefficients and genealogical relationships were then calculated from the multilocus genotypes with the software package ML-Relate (Kalinowski et al. 2006). These estimates were used to infer whether kinship clusters would mirror the social structure revealed by network visualization here described as communities and cliques (figure 1).

As it was not our aim to detect the most trivial kinship structure given by mother–offspring dyads, all known mother–offspring pairs where the offspring was still nursed were excluded. Mother–offspring pairs were known with high confidence (see Trillmich & Wolf 2008, see also cross fostering below). To further rule out that kinship structure was determined by offspring that had been weaned only recently, all analyses were carried out on a subset of the network restricted to adult females. All statistical inference on kinship structure was based on randomization techniques that were programed in R (R Development Core Team 2006). We deliberately applied three different approaches to validate the robustness of the results. In the following, we outline the general logic behind each approach (for further details, see electronic supplementary material).

(i) Matrix correlation between social tie strength and relatedness coefficient

The network in figure 1 is a binary visualization of a symmetric matrix that summarizes social association of individuals in local social groups over 141 behavioural surveys. Pairs of individuals that have been sighted at least twice together in a local group during all surveys are inter-connected (for details on how social association based on shared group membership is defined, see Wolf et al. 2007a). Contrary to the impression of the binary network (figure 1), the total number of times that a pair was found together in a local group (tie strength) varies considerably and ranges from 0 to 25. Tie strength thus provides a quantitative measure for social affinity and can be used to test whether social bonds between related individuals are stronger than between less related individuals. This is achieved by comparing the tie strength matrix to the matrix containing the estimates of pairwise relatedness using a partial Mantel test (ecodist package in R, 10 000 permutations). To prevent the influence of spurious observations, pairs that were seen together less than twice were excluded.

(ii) Relatedness coefficients and simulated Wilcoxon test statistic

The general idea behind this approach is to compare the degree of average pairwise relatedness between the different hierarchical components of the network: cliques; communities; and the global population background. To this end, all possible relationships between individuals in the network were divided into two groups: pairs of individuals that share membership of a given unit and those that come from different units (e.g. same community versus different communities). Groups were then statistically compared using a randomized Wilcoxon rank sum test with 10 000 Monte Carlo resamplings. Relative frequencies of simulated values that are equal to or larger than the empirical value can be interpreted as the p-value, the probability of committing a type I error (Manly 1997).

(iii) Pedigrees, simulated contingency tables and Χ2-test statistic

Along the lines of the aforementioned procedure, this approach compares relatedness coefficients via an estimated relationship category between individuals. Relationship categories include parent–offspring (PO), full sibling (FS), half sibling (HS) and unrelated (U), as estimated on the basis of shared alleles. Analysis was restricted to individual pairs where the relationship was attributed with at least 95% confidence based on 5000 iterations (Kalinowski et al. 2006); ambiguous categories such as HSorFS, FSorPO, etc. were excluded. All pairs where specific estimates were available were again divided into two groups. Frequency information of each relationship category was then summarized in a contingency table for both groups and a randomized Χ2-statistic with 10 000 Monte Carlo resamplings was used to test for statistical independence between groups. Simulated Χ2 residuals were used to infer which of the categories significantly contributed to the overall effect.

(b) Cross fostering

Galápagos sea lions give birth to only one pup every year or every second year (Trillmich & Wolf 2008). Cross-fostering events can therefore be determined easily by a combination of behavioural monitoring of suckling bouts and genetic relatedness analyses. From 2003 to 2005, we recorded suckling bouts between individually tagged female–young pairs during a total of 510 behavioural surveys (Wolf & Trillmich 2007). To eliminate misidentifications by tag reading errors, we restricted the analyses to suckling pairs that had been sighted at least three times or had been explicitly confirmed by the observer. Altogether, a total of 2304 suckling bouts were observed in 176 pairs of individually identified female–young pairs. These pairs included 123 females that were seen with different offspring over the years and 174 young, two of which were seen suckling multiple females. After genotyping all pairs we used Cervus v. 3.0 (Marshall et al. 1998; Kalinowski et al. 2007) to determine whether the observed suckling pairs were biological mother–offspring units. The most likely mother was estimated for each offspring from the entire candidate set of females with a 99% confidence threshold. Pairs with no significant match or where the inferred genetic mother–offspring combinations differed from the suckling pairs observed in the field were scored as cross-foster pairs. The degree of genetic relatedness between these pairs was estimated with ML-Relate (Kalinowski et al. 2006).

3. Results

(a) Social network analyses

(i) Kinship structure is tightly linked to spatial behaviour

Overall relatedness was low in both cliques and communities (table 1). Degree of relatedness was higher among individuals within than among communities. Average within-community relatedness was twice as high as among communities (Wilcoxon statistic: p<0.001, Χ2-statistic: p=0.007; table 1). This pattern held if only adult females were considered (115 females instead of 379 individuals of the overall network). Adult females sharing community membership were on average 1.82 times more related than females of different communities (Wilcoxon statistic: p=0.051, Χ2-statistic: p=0.011; table 1). In both cases (overall network and females only), Χ2 residuals showed that the number of full siblings, half siblings and former mother–offspring pairs was higher within communities than between them. Unrelated individuals were found significantly more often across different communities (figure 2).

Table 1.

Average relatedness between individuals in the sea lion social network (figure 1). We report average relatedness between individuals that share membership of a given tier (within) or that are associated with different tiers (between). The ratio (within/between) indicates, if within tier relatedness is higher (more than 1) or lower (less than 1) than between tiers. Statistical significance is symbolized by asterisks according to common standards; n.s., not significant.

mean relatedness median relatedness
all individuals cliques within 0.0505 0.0065
between 0.0499 0.008
within/betweenn.s. 1.0 0.80
communities within 0.0500 0.0075
between 0.0469 0.0037
within/between*** 1.07 2.02
adult females cliques within 0.0554 0.00951
between 0.0522 0.0073
within/betweenn.s. 1.06 1.30
communities within 0.053 0.008
between 0.047 0.004
within/between* 1.12 1.82
Open in a new tab
Figure 2.

Figure 2

Open in a new tab

Relative contribution of different pairwise relationship categories to kinship structure in social communities of the Galapagos sea lion (figure 1a). Individuals from the same community (white bars) are compared with those from different communities (grey bars). Depicted are simulated residual values that show how strong a given category deviates from the random expectation of no difference between the two groups (i.e. no kinship structure). Positive values indicate that a category is relatively overrepresented, negative values that it is underrepresented. Asterisks indicate how well the deviation is statistically supported. (a) Comparison including all 379 individuals of the entire network. (b) Comparison restricted to 115 adult females. Category abbreviations: HS, half sibling; FS, full sibling; PO, parent–offspring.

(ii) Targeted kin selection between highly related individuals does not create kin clusters

If social relationships in the colony were mainly driven by targeted kin selection, we would expect social cliques to be the core units of relatedness. Under this prediction, the median degree of pairwise relatedness within a clique should be significantly higher than between cliques of the same community. However, the degree of relatedness within cliques was not distinguishable from the community background (simulated Wilcoxon statistic: p=0.581, simulated Χ2-statistic: p=0.410). Nonetheless, when observed mother–offspring pairs were included in the analysis, median relationship within cliques exceeded that of the community background (Wilcoxon statistic: p=0.002, Χ2-statistic: p<0.001). Even so, animals within cliques were only 1.06 times more closely related than between cliques. From this, we conclude that in the social core unit of the network (cliques) kin selection only acts in its most trivial form of direct parental care.

(iii) Social tie strength and genetic relatedness are correlated

Strength of social relationships was approximated by the number of times that a pair of individuals was sighted together (tie strength; see §2). Thus, we can ask whether behavioural decisions at the level of interaction were influenced by the degree of relatedness. Indeed, we found that the tie strength and the degree of relatedness between connected pairs of individuals were correlated (Mantel test: r=0.16, p<0.001; figure 3).

Figure 3.

Figure 3

Open in a new tab

Correlation of minimum tie strength and relatedness between pairs of individuals that are connected in the social network (figure 1a). A non-parametric lowess smoothing line is shown for illustration. Total number of connections between individuals in the network with a given minimum tie strength is shown at the bottom.

(b) Cross fostering

During three years of observation, we recorded a cross-foster frequency of 0.006% (14 non-filial/2304 total suckling events). Parentage was genetically inferred for 97% (171 out of 176) of suckling female–pup pairs. Of the five pairs where pup and female did not fit genetically, only one case could be scored as true non-filial nursing (less than 0.01% of all pairs). The remaining four pairs included one adult male (greater than female size) that was seen suckling once, and three young that usually suckled their biological mother and were only observed to steal milk occasionally. Non-filial nursing, therefore, did not represent stable cross-fostering relationships. Overall relatedness of cross-suckling pairs was very similar to the overall relatedness background of random pairs in the immediate social neighbourhood (rmean=0.086, rmedian=0.06, compare with table 1).

4. Discussion

Using network analyses, we investigated the relationship between social and genetic structures and revisited the potential role of kin selection for the occurrence of altruistic behaviour in a colonial mammal. Genetic structure—inferred by both genetic relatedness coefficients and pedigree relationships—coincided with ‘social communities’, an intermediate tier of the social network that closely corresponds to spatial relationships among individuals. This finding supports the idea that kin structure can arise as an emergent property of limited dispersal alone and need not imply any targeted, beneficial interactions among kin. These results are consistent with many studies showing that philopatry is a powerful mechanism by which genetic structure can accumulate as a by-product (Pomeroy et al. 2000; Fowler 2005; Fabiani et al. 2006; Campbell et al. 2008). This makes intuitive sense in the Galápagos sea lion, a species where female offspring are prone to inherit their mother's home range (Wolf & Trillmich 2007) and where male site fidelity might be related to long-term dominance hierarchies.

While spatial assortment bore on genetic structure, genetic relatedness among the lower level tiers in the network (cliques), which are defined by social preference, was not elevated. Likewise, altruistic cross-fostering behaviour was virtually absent and the exceptional female–young dyads that displayed cross-fostering showed no extreme degree of relatedness. Cross-fostering events appear to be best explained as occasional cases of misdirected parental care as has been suggested for other species (reviewed by Roulin 2002). Similarly, all previous studies that looked into the degree of relatedness between offspring and their foster mothers did not find any evidence for kin selection (Perry et al. 1998; McCulloch et al. 1999; Schaeff et al. 1999; Hoffman & Amos 2005). These results highlight how difficult it can be to interpret kinship structure in animal societies. Although it is often tempting to infer inclusive fitness benefits from correlational relationships between kinship structure and social behaviour, our findings point to the fact that co-operative behaviour need not evolve despite kinship structure, or may well be rooted in other non-genetic mechanisms (Clutton-Brock 2002).

This result notwithstanding, even a slightly elevated background of relatedness holds the potential for weakly selected co-operative behaviours to evolve. Compared with dynasties of communally breeding species, the average degree of relatedness in large animal groups should be low. Nonetheless, preferential associations among genetically more similar individuals may still be advantageous. This idea becomes particularly plausible if interactions are iterated many times between individuals with long-term memory (for sea lions, see Kastak & Schusterman 2002). If costs to the donor are small compared with the benefits gained by the receiver, low levels of average relatedness can suffice to induce co-operative behaviour.

Our finding that individuals maintained stronger social ties with genetically more similar interaction partners indeed suggests that the degree of relatedness, though comparatively low overall, influences behavioural decisions. Taking into account that a considerable amount of noise is artificially introduced by estimating low levels of relatedness with only 22 microsatellite markers (Blouin et al. 1996), we can assume that the correlation is stronger than it appears. Technically, this result is important insofar as it suggests that networks weighted by some additional criterion may outperform strictly binary networks (Whitehead 2008). More essentially, it raises the fascinating prospect that low levels of relatedness and even small differences in relatedness can be relevant for natural selection to act on certain types of social behaviour other than the sensational text book examples of cross fostering or ‘suicidal sacrifice in the name of others’ (Alcock 2005). We know that kin recognition works even with intriguingly low differences in relatedness (Jehle et al. 2007) and honest olfactory cues signalling an individual's immunogenetic composition probably exist (Boehm & Zufall 2006). We further know that animals are able to adjust their degree of aggressiveness to the degree of relatedness of their opponent (Hanggi & Schusterman 1990; Piertney et al. 1999) and that this can pay in terms of fitness. Stress induced by aggressive behaviour not only towards non-filial young but also towards adult females is known to negatively impact on survival and fitness (Le Boeuf & Briggs 1977; Vilá & Cassini 1990; Ceacero et al. 2007). This predicts a role of less obvious co-operative behaviours such as social tolerance for the successful formation and stability of animal aggregations (Hare et al. 2007).

Acknowledgments

We thank our numerous field assistants for their excellent help in data collection and cheerful companionship on a small islet and gratefully acknowledge the support of Diethard Tautz in whose laboratory the genetic work was conducted. We further thank Richard James for stimulating discussions and revising an earlier draft of this manuscript. We are grateful to two anonymous referees and to Theodore G. Manno who helped streamlining the manuscript. Karl Lutzow and Begon Bloodhurst inspired the work with their asymmetric sophism. The study was financed through the Deutsche Forschungsgemeinschaft (WO 1426/1-1) and supported by the Max Planck Society through the Institute for Ornithology, Seewiesen. We would like to thank the Galápagos National Park for the permit to work with and take samples of sea lions. The Charles Darwin Station provided valuable logistic support.

Supplementary Material

Detailed methods

Detailed description of randomization techniques used to test for statistical association of genetic relatedness and network structure

rspb20080356s01.doc (39.5KB, doc)

References

  1. Alcock J. Sinauer Associates, Inc; Sunderland, MA: 2005. Animal behavior: an evolutionary approach. [Google Scholar]
  2. Blouin M.S, Parsons M, Lacaille V, Lotz S. Use of microsatellite loci to classify individuals by relatedness. Mol. Ecol. 1996;5:393–401. doi: 10.1111/j.1365-294x.1996.tb00329.x. doi:10.1111/j.1365-294X.1996.tb00329.x [DOI] [PubMed] [Google Scholar]
  3. Boehm T, Zufall F. MHC peptides and the sensory evaluation of genotype. Trends Neurosci. 2006;29:100–107. doi: 10.1016/j.tins.2005.11.006. doi:10.1016/j.tins.2005.11.006 [DOI] [PubMed] [Google Scholar]
  4. Boness D.J, Craig M.P, Honigman L, Austin S. Fostering behavior and the effect of female density in Hawaiian monk seals, Monachus schauinslandi. J. Mammal. 1998;79:1060–1069. doi:10.2307/1383115 [Google Scholar]
  5. Campbell R.A, Gales N.J, Lento G.M, Baker C.S. Islands in the sea: extreme female natal site fidelity in the Australian sea lion, Neophoca cinerea. Biol. Lett. 2008;4:139–142. doi: 10.1098/rsbl.2007.0487. doi:10.1098/rsbl.2007.0487 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Ceacero F, Landete-Castillejos T, Garcia A.J, Estevez J.A, Gallego L. Kinship discrimination and effects on social rank and aggressiveness levels in Iberian red deer hinds. Ethology. 2007;113:1133–1140. doi:10.1111/j.1439-0310.2007.01427.x [Google Scholar]
  7. Childerhouse S, Gales N. Fostering behaviour in New Zealand sea lions Phocarctos hookeri. N Z J. Zool. 2001;28:189–195. [Google Scholar]
  8. Clutton-Brock T. Behavioral ecology—breeding together: kin selection and mutualism in cooperative vertebrates. Science. 2002;296:69–72. doi: 10.1126/science.296.5565.69. doi:10.1126/science.296.5565.69 [DOI] [PubMed] [Google Scholar]
  9. Croft D.P, James R, Thomas P.O.R, Hathaway C, Mawdsley D, Laland K.N, Krause J. Social structure and co-operative interactions in a wild population of guppies (Poecilia reticulata) Behav. Ecol. Sociobiol. 2006;59:644–650. doi:10.1007/s00265-005-0091-y [Google Scholar]
  10. Dobson F.S. Competition for mates and predominant juvenile male dispersal in mammals. Anim. Behav. 1982;30:1183–1192. doi:10.1016/S0003-3472(82)80209-1 [Google Scholar]
  11. Dobson F.S. Social structure and gene dynamics in mammals. J. Mammal. 1998;79:667–670. [Google Scholar]
  12. Fabiani A, Galimberti F, Sanvito S, Hoelzel A.R. Relatedness and site fidelity at the southern elephant seal, Mirounga leonina, breeding colony in the Falkland Islands. Anim. Behav. 2006;72:617–626. doi:10.1016/j.anbehav.2005.11.024 [Google Scholar]
  13. Faulkes C.G, Abbott D.H, O'brien H.P, Lau L, Roy M.R, Wayne R.K, Bruford M.W. Micro- and macrogeographical genetic structure of colonies of naked mole-rats Heterocephalus glaber. Mol. Ecol. 1997;6:615–628. doi: 10.1046/j.1365-294x.1997.00227.x. doi:10.1046/j.1365-294X.1997.00227.x [DOI] [PubMed] [Google Scholar]
  14. Fowler A.C. Fine-scale spatial structuring in cackling Canada geese related to reproductive performance and breeding philopatry. Anim. Behav. 2005;69:973–981. doi:10.1016/j.anbehav.2004.07.014 [Google Scholar]
  15. Griffin A.S, West S.A. Kin discrimination and the benefit of helping in cooperatively breeding vertebrates. Science. 2003;302:634–636. doi: 10.1126/science.1089402. doi:10.1126/science.1089402 [DOI] [PubMed] [Google Scholar]
  16. Hamilton W.D. The genetical evolution of social behaviour I. J. Theor. Biol. 1964;7:1–16. doi: 10.1016/0022-5193(64)90038-4. doi:10.1016/0022-5193(64)90038-4 [DOI] [PubMed] [Google Scholar]
  17. Hanggi E.B, Schusterman R.J. Kin recognition in captive Californian sea lions (Zalophus californianus) J. Comp. Psychol. 1990;104:368–372. doi:10.1037/0735-7036.104.4.368 [Google Scholar]
  18. Hare B, Melis A.P, Woods V, Hastings S, Wrangham R. Tolerance allows bonobos to outperform chimpanzees on a cooperative task. Curr. Biol. 2007;17:619–623. doi: 10.1016/j.cub.2007.02.040. doi:10.1016/j.cub.2007.02.040 [DOI] [PubMed] [Google Scholar]
  19. Hoffman J.I, Amos W. Does kin selection influence fostering behaviour in Antarctic fur seals (Arctocephalus gazella)? Proc. R. Soc. B. 2005;272:2017–2022. doi: 10.1098/rspb.2005.3176. doi:10.1098/rspb.2005.3176 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hoffman J.I, Steinfartz S, Wolf J.B.W. Ten novel dinucleotide microsatellite loci cloned from the Galápagos sea lion (Zalophus californianus wollebaeki) are polymorphic in other pinniped species. Mol. Ecol. Notes. 2007;7:103–105. doi:10.1111/j.1471-8286.2006.01544.x [Google Scholar]
  21. Jehle R, Sztatecsny M, Wolf J.B.W, Whitlock A, Hödl W, Burke T. Genetic dissimilarity predicts paternity in the smooth newt (Lissotriton vulgaris) Biol. Lett. 2007;3:526–528. doi: 10.1098/rsbl.2007.0311. doi:10.1098/rsbl.2007.0311 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kalinowski S.T, Wagner A.P, Taper M.L. ML-Relate: a computer program for maximum likelihood estimation of relatedness and relationship. Mol. Ecol. Notes. 2006;6:576–579. doi:10.1111/j.1471-8286.2006.01256.x [Google Scholar]
  23. Kalinowski S.T, Taper M.L, Marshall T.C. Revising how the computer program Cervus accomodates genotyping error increases success in paternity assignment. Mol. Ecol. 2007;16:1099–1106. doi: 10.1111/j.1365-294X.2007.03089.x. doi:10.1111/j.1365-294X.2007.03089.x [DOI] [PubMed] [Google Scholar]
  24. Kastak C.R, Schusterman R.J. Long-term memory for concepts in a California sea lion (Zalophus californianus) Anim. Cogn. 2002;5:225–232. doi: 10.1007/s10071-002-0153-8. doi:10.1007/s10071-002-0153-8 [DOI] [PubMed] [Google Scholar]
  25. Krause J, Croft D.P, James R. Social network theory in the behavioural sciences: potential applications. Behav. Ecol. Sociobiol. 2007;62:15–27. doi: 10.1007/s00265-007-0445-8. doi:10.1007/s00265-007-0445-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Le Boeuf B.J, Briggs K.T. The cost of living in a seal harem. Mammalia. 1977;41:167–195. [Google Scholar]
  27. Lehmann L, Keller L, Sumpter D.J.T. The evolution of helping and harming on graphs: the return of the inclusive fitness effect. J. Evol. Biol. 2007;20:2284–2295. doi: 10.1111/j.1420-9101.2007.01414.x. doi:10.1111/j.1420-9101.2007.01414.x [DOI] [PubMed] [Google Scholar]
  28. Lion S, van Baalen M. Self-structuring in spatial evolutionary ecology. Ecol. Lett. 2008;11:277–295. doi: 10.1111/j.1461-0248.2007.01132.x. doi:10.1111/j.1461-0248.2007.01132.x [DOI] [PubMed] [Google Scholar]
  29. Lukas D, Reynolds V, Boesch C, Vigilant L. To what extent does living in a group mean living with kin? Mol. Ecol. 2005;14:2181–2196. doi: 10.1111/j.1365-294X.2005.02560.x. doi:10.1111/j.1365-294X.2005.02560.x [DOI] [PubMed] [Google Scholar]
  30. Lusseau D, Wilson B, Hammond P.S, Grellier K, Durban J.W, Parsons K.M, Barton T.R, Thompson P.M. Quantifying the influence of sociality on population structure in bottlenose dolphins. J. Anim. Ecol. 2005;75:14–24. doi: 10.1111/j.1365-2656.2005.01013.x. doi:10.1111/j.1365-2656.2005.01013.x [DOI] [PubMed] [Google Scholar]
  31. Manly B.F.J. Chapman & Hall; London, UK: 1997. Randomization, bootstrap and Monte Carlo methods in biology. [Google Scholar]
  32. Marshall T.C, Slate J, Kruuk L.E.B, Pemberton J.M. Statistical confidence for likelihood-based paternity inference in natural populations. Mol. Ecol. 1998;7:639–655. doi: 10.1046/j.1365-294x.1998.00374.x. doi:10.1046/j.1365-294x.1998.00374.x [DOI] [PubMed] [Google Scholar]
  33. McCulloch S, Pomeroy P.P, Slater P.J.B. Individually distinctive pup vocalizations fail to prevent allo-suckling in grey seals. Can. J. Zool. 1999;77:716–723. doi:10.1139/cjz-77-5-716 [Google Scholar]
  34. Ohtsuki H, Hauert C, Lieberman E, Nowak M.A. A simple rule for the evolution of cooperation on graphs and social networks. Nature. 2006;441:502–505. doi: 10.1038/nature04605. doi:10.1038/nature04605 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Perry E.A, Boness D.J, Fleischer R.C. DNA fingerprinting evidence of nonfilial nursing in grey seals. Mol. Ecol. 1998;7:81–85. doi: 10.1046/j.1365-294x.1998.00313.x. doi:10.1046/j.1365-294x.1998.00313.x [DOI] [PubMed] [Google Scholar]
  36. Piertney S.B, MacColl A.D.C, Lambin X, Moss R, Dallas J.F. Spatial distribution of genetic relatedness in a moorland population of red grouse (Lagopus lagopus scoticus) Biol. J. Linn. Soc. 1999;68:317–331. [Google Scholar]
  37. Pomeroy P.P, Twiss S.D, Redman P. Philopatry, site fidelity and local kin associations within grey seal breeding colonies. Ethology. 2000;106:899–919. doi:10.1046/j.1439-0310.2000.00610.x [Google Scholar]
  38. R Development Core Team. R Foundation for Statistical Computing; Vienna, Austria: 2006. R: a language and environment for statistical computing. [Google Scholar]
  39. Riedman M.L, Le Boeuf B.J. Mother-pup separation and adoption in northern elephant seals. Behav. Ecol. Sociobiol. 1982;11:203–215. doi:10.1007/BF00300063 [Google Scholar]
  40. Roulin A. Why do lactating females nurse alien offspring? A review of hypotheses and empirical evidence. Anim. Behav. 2002;63:201–208. doi:10.1006/anbe.2001.1895 [Google Scholar]
  41. Schaeff C.M, Boness D.J, Bowen W.D. Female distribution, genetic relatedness, and fostering behaviour in harbour seals, Phoca vitulina. Anim. Behav. 1999;57:427–434. doi: 10.1006/anbe.1998.1001. doi:10.1006/anbe.1998.1001 [DOI] [PubMed] [Google Scholar]
  42. Sugg D.W, Chesser R.K, Dobson F.S, Hoogland J.L. Population genetics meets behavioral ecology. Trends Ecol. Evol. 1996;11:338–342. doi: 10.1016/0169-5347(96)20050-3. doi:10.1016/0169-5347(96)20050-3 [DOI] [PubMed] [Google Scholar]
  43. Taylor P.D, Day T, Wild G. Evolution of cooperation in a finite homogeneous graph. Nature. 2007;447:469–472. doi: 10.1038/nature05784. doi:10.1038/nature05784 [DOI] [PubMed] [Google Scholar]
  44. Trillmich F, Wolf J.B.W. Parent–offspring and sibling conflict in Galápagos fur seals and sea lions. Behav. Ecol. Sociobiol. 2008;62:363–375. doi:10.1007/s00265-007-0423-1 [Google Scholar]
  45. Vilá B.L, Cassini M.H. Agresividad entre hembras y separación madre-cria en el lobo marino del sur, en Chubut, Argentina. Revista Chilena Hist. Nat. 1990;63:169–176. [Google Scholar]
  46. Wey T, Blumstein D.T, Shen W, Jordán F. Social network analysis of animal behaviour: a promising tool for the study of sociality. Anim. Behav. 2008;75:333–344. doi:10.1016/j.anbehav.2007.06.020 [Google Scholar]
  47. Whitehead H. University of Chicago Press; Chicago, IL: 2008. Analyzing animal societies: quantitative methods for vertebrate social analysis. [Google Scholar]
  48. Wolf J.B.W, Trillmich F. Beyond habitat requirements: individual fine-scale site fidelity in a colony of the Galapagos sea lion (Zalophus wollebaeki) creates conditions for social structuring. Oecologia. 2007;152:553–567. doi: 10.1007/s00442-007-0665-7. doi:10.1007/s00442-007-0665-7 [DOI] [PubMed] [Google Scholar]
  49. Wolf J.B.W, Kauermann G, Trillmich F. Males in the shade: habitat use and sexual segregation in the Galápagos sea lion (Zalophus californianus wollebaeki) Behav. Ecol. Sociobiol. 2005;59:293–302. doi:10.1007/s00265-005-0042-7 [Google Scholar]
  50. Wolf J.B.W, Tautz D, Caccone A, Steinfartz S. Development of new microsatellite loci and evaluation of loci from other pinniped species for the Galápagos sea lion (Zalophus californianus wollebaeki) Conserv. Genet. 2006;7:461–465. doi:10.1007/s10592-005-9045-1 [Google Scholar]
  51. Wolf J.B.W, Mawdsley D, Trillmich F, James R. Social structure in a colonial mammal: unravelling hidden structural layers and their foundations by network analysis. Anim. Behav. 2007a;74:1293–1302. doi:10.1016/j.anbehav.2007.02.024 [Google Scholar]
  52. Wolf J.B.W, Tautz D, Trillmich F. Galápagos and Californian sea lions are separate species: genetic analysis of the genus Zalophus and its implications for conservation management. Front. Zool. 2007b;4:1–13. doi: 10.1186/1742-9994-4-20. doi:10.1186/1742-9994-4-20 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Wolf J.B.W, Harrod C, Brunner S, Trillmich F, Tautz D. Tracing early stages of species differentiation: ecological, morphological and genetic divergence of Galápagos sea lion populations. Bmc Evol. Biol. 2008;8:150. doi: 10.1186/1471-2148-8-150. doi:10.1186/1471-2148-8-150 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Detailed methods

Detailed description of randomization techniques used to test for statistical association of genetic relatedness and network structure

rspb20080356s01.doc (39.5KB, doc)

Articles from Proceedings of the Royal Society B: Biological Sciences are provided here courtesy of The Royal Society

RESOURCES