Behav Ecol Sociobiol (2009) 63:967–973

DOI 10.1007/s00265-009-0747-0

REVIEW

Animal social networks: an introduction

Jens Krause & David Lusseau & Richard James

Received: 2 January 2009 / Revised: 2 March 2009 / Accepted: 2 March 2009 / Published online: 2 April 2009
# Springer-Verlag 2009

Abstract Network analysis has a long history in the heterogeneous relationships in a population. Probing the
mathematical and social sciences and the aim of this biological drivers for these heterogeneities, often as a
introduction is to provide a brief overview of the potential function of time, forms the basis of many of the current
that it holds for the study of animal behaviour. One of the uses of network analysis in the behavioural sciences. This
most attractive features of the network paradigm is that it special issue on social networks brings together a diverse
provides a single conceptual framework with which we can group of practitioners whose study systems range from
study the social organisation of animals at all levels social insects over reptiles to birds, cetaceans, ungulates
(individual, dyad, group, population) and for all types of and primates in order to illustrate the wide-ranging
interaction (aggressive, cooperative, sexual etc.). Graphical applications of network analysis.
tools allow a visual inspection of networks which often
helps inspire ideas for testable hypotheses. Network Keywords Social networks . Animal interactions .
analysis itself provides a multitude of novel statistical tools Social organisation
that can be used to characterise social patterns in animal
populations. Among the important insights that networks
have facilitated is that indirect social connections matter. Introduction
Interactions between individuals generate a social environ-
ment at the population level which in turn selects for At the mention of social networks, many may think “So
behavioural strategies at the individual level. A social what, hasn’t this method been around for decades?” and
network is often a perfect means by which to represent they are right. The origins of the network concept are in
mathematical graph theory that can be traced back for
Communicated by T. Czeschlik centuries (Euler 1736 in Biggs et al. 1986) and its use in
social network analysis was instigated by sociologists and
This contribution is part of the special issue “Social Networks: new
perspectives” (Guest Editors: J. Krause, D. Lusseau and R. James). psychologists in the first half of the 20th century (Scott
2000). However, what’s new is the fact that the subject has
J. Krause (*)
undergone a dramatic and rapid transformation in
Institute of Integrative and Comparative Biology,
University of Leeds, recent years. Increasing globalisation has seemingly
Leeds LS2 9JT, UK interconnected everything, our communication networks
e-mail: [email protected] (e.g. internet, mobile phone) and transport systems span the
globe, economic integration of nations and communities
D. Lusseau
Institute of Biological and Environmental Sciences, becomes ever closer and more complex with a growing
University of Aberdeen, number of multinational organisations and environmental
Aberdeen AB24 2TZ, UK problems transcending national boundaries. The fact that
our world is becoming more and more interconnected has
R. James
Department of Physics, University of Bath, led to an exponential interest in understanding networks
Bath BA2 7AY, UK because they are an effective way of studying this process
968 Behav Ecol Sociobiol (2009) 63:967–973

of increasing social inter-connectedness (see Newman 2003 social (network) structure of the population can have
for a review). As a result of this strong interest, there have important repercussions for the fitness of individuals
been major conceptual advances in several areas, including (Fig. 2). This argument is familiar to most of us from
statistical physics that have pushed the boundaries of our game theoretic models. The frequency at which different
understanding of networked systems. Software programs behavioural strategies are used in a population can have
have been developed that make the use of network analyses important fitness consequences for individuals in the
more user-friendly, opening these methods up for the population (Maynard Smith 1982). Early game theoretic
mainstream. So far, however, the fields of animal behaviour models assumed that all individuals mix freely with each
and behavioural ecology have benefited relatively little other. In most animal and human populations, however, not
from the new possibilities that these recent analytical everybody interacts with everybody else and we see a
advances offer us and the purpose of this special edition highly structured social organisation that reflects differ-
is to address this issue. ences between individuals in the number of social inter-
So what’s so special about network analyses then? actions, the degree to which some individuals are central or
Network analysis provides tools which allow us to test for peripheral to the population network and the tendency to
the first time some of the long-standing conceptual frame- interconnect different communities that form substructures
works of social organisation and structure (e.g. Wilson within networks (Krause and Ruxton 2002; Krause et al.
1975; Hinde 1976). Most conventional analyses might 2007; Croft et al. 2008). More recent “evolutionary graph
permit exploration of social structure at the level of the theory” models (for example Ohtsuki et al. 2006; Santos et
individual, dyad, group or population. Network analyses al. 2008) use networks to quantify social heterogeneity, and
allow us to span these scales, and scales in between, account for it, in models of, for example, the evolution and
making it possible for us to build up larger social units from maintenance of cooperation. While the nature of the “social
pair-wise interactions. Though such analyses have long networks” invoked in evolutionary game theory may be
been possible for some types of relational data (such as rather different from the social networks observed and
associations among group-living animals), the strength of analysed by empiricists, the parallel is an interesting one.
the network approach is that so many types of interaction A similar point can be made about conventional
(sexual, aggressive, cooperative; who eats whom and so on) epidemiological models and the modelling of socially
can be treated within the same conceptual framework, and transmitted information. The initial assumption of random
using the same visual and analytic tools. Perhaps “Animal interactions between individuals turns out to be too
Interaction Networks” would be a more suitable title for simplistic and does not fit the finding that most social
this field, given the generality of the potential applications. systems have a more or less heterogeneous structure that
This generality means we can look at how individual should not be neglected when studying the processes that
behaviour influences what happens at the population level and take place on these systems.
we can likewise study the fitness implications of the latter for
the individual (Fig. 1). This kind of feedback loop is
fundamental for an understanding of the role of self- An example
organisation in social systems (Camazine et al. 2001; Couzin
and Krause 2003). A great advantage of a networks approach A network can be described as consisting of nodes
is that we do not need to decide a priori on the level of social (individuals) and edges (interactions between them)
organisation that must be key to understanding our animal (Fig. 3). Part of the appeal of the network approach is its
system. We always deal simply with nodes (animals) and
edges (interactions between them) and search the patterns
they reveal, through statistical tools, models or both. Behaviour depends
on what others do
Finally, the network approach also allows the contem-
plation of a much neglected aspect: specifically that the

influences Individual How many

others are there?
Individual Population
behaviour dynamics Who is connected
to whom?
selects for
Fig. 2 Schematic representation of the importance of the frequency of
Fig. 1 Schematic relationship between individual behaviour and behavioural strategies and the social fine-structure for the evolution of
population dynamics (after Kokko unpublished) behaviour
Behav Ecol Sociobiol (2009) 63:967–973 969

simplicity and generality because almost any system that Table 1 Some individual-based measures for the network in Fig. 3
comprises multiple components (whether biological or Node/ Node Path length Clustering Node
technological) can be described in the form of a network. individual degree (to/from coefficient betweenness
In Fig. 3, we have constructed a simple example to a node)
illustrate some of the features that network theory offers.
a 1 3.64 0 0
We can calculate a whole new range of descriptive statistics
b 4 2.73 0.333 10.5
that can be used to characterise structural components of
c 3 2.36 0.333 5.8
the networks and the position of chosen individuals in
d 3 2.36 0.667 4.3
relation to others. For example, we can calculate for each
individual in the network its degree (number of immediate e 4 2.27 0.5 5.3
neighbours), cluster coefficient (the degree to which an f 3 2.09 0.333 8.8
individual’s immediate neighbours are connected), path g 4 2 0.333 18.2
length (number of connections on the shortest path between h 3 2 0.333 28
two individuals) and node betweenness (the number of i 4 2.36 0.333 24.5
shortest paths between pairs of individuals that pass j 2 3.18 1 0
through a particular individual) (see Croft et al. 2008 for k 2 3.18 1 0
details). These statistics (which are just a small proportion l 3 3.09 0.667 0.5
of those already available from the social sciences Mean values 3 2.61 0.486 8.83
literature) can be averaged over all individuals in the
See text for a definition of the network measures. In the case of the
network to give an idea of the local and global properties of path length, we have calculated the mean distance of nodes to and
the network (Table 1). from a particular node
Of course, the measures we use to characterise network
structure are most useful if they reflect structures of
biological interest. An individual’s degree, for example, readily available from the internet. Other measures, such as
might be related to the likelihood that the individual can reach and, in particular, measures of “centrality” are now
spread disease or information in a network, with highly being developed and used by behavioural scientists, as
connected individuals more likely to trigger an epidemic or readers of the papers in this volume will see for themselves.
a rumour. Path length and cluster coefficient might tell us Lusseau et al. (2008) commented on which measures may
something about the likelihood that a pathogen will remain be particularly appropriate for animal behaviour data.
a local outbreak or become global in a population.
Betweenness may indicate how important individuals are
in interconnecting different sections of the network. In the Some of the uses of network theory
context of social learning, individuals with high between-
ness that can reach into different communities may be more Our hope is that the papers in this edition will provide a
likely to be responsible for the global spread of informa- whole range of examples of and ideas for the use of
tion. Many of these network statistics and others can be networks in the behavioural sciences that will inspire
calculated using the social sciences package UCINET readers to become active themselves in this exciting and
(Borgatti et al. 2002; see also Croft et al. 2008) which is rapidly developing research area.
In many ways, the developments in the use of network
analysis in the behavioural sciences have tracked similar
c developments in the social sciences. We have the advan-
f
tage, though, in that “discovering” networks several
j decades later than social scientists we have ready access
b
e h
to their ideas and methods, plus those that continue to
i
l
emerge in other fields. We also have cheap computing
a
power to implement sometimes rather involved statistical
g
tests and analyses.
k It is probably fair to say that the majority of applications
d of network analysis in the behavioural sciences to date have
been of a descriptive nature; a network is constructed and
Fig. 3 Example of a social network where nodes (black circles)
analysed for the patterns it contains. Most of us have
symbolise individuals and edges (lines) social connections between
them. This fictitious network comprises 12 individuals (labelled a–l). employed what Whitehead and Dufault (1999) called the
See Table 1 for individual-based measures “gambit of the group”, in which individuals are assumed to
970 Behav Ecol Sociobiol (2009) 63:967–973

be associating if they are seen in the same group; these sampling. They were able to conclude that bachelors form
associations are accumulated over a number of observa- stronger, more persistent bonds than stallions.
tions, to produce an association matrix, which is itself a This last example highlights that relational data, such as a
representation of a social network. Thus, all the tools network, need particular care at the analysis stage, since we can
developed for the analysis of association data (Whitehead rarely assume independence of our data points. Though there
2008) may be regarded as tools available for network are plenty of methods already available (Whitehead 2008;
analysis of data compiled via the gambit of the group, even Croft et al 2008), new questions still often require at least a
if they are not always explicitly reported as such. tweak on old methods, or sometimes a completely new
This approach, whether explicitly using networks or not, approach. There is still plenty of room for developments in
can help us identify patterns of social organisation network methodology, many of which appear in the various
(including how they change with time) that can lay the contributions to this volume. James et al. (2009) warn us of
foundation for understanding key components of social some of the potential pitfalls of applying some of the existing
structure, and for making comparisons of different pop- methods, analyses and interpretations without quite enough
ulations, contexts or species. This approach has plenty of thought. Krause et al. (2009) point out that, even for those of
mileage left in it. Early network examples include the work us that have appreciated the need for careful statistical
of Lusseau (2003) on dolphins and Croft et al. (2004) on analysis, it is still surprisingly easy to build a null model of
guppies, Poecilia reticulata. In this volume, there are many social associations that is itself not without bias. On a slightly
excellent examples of the use and analysis of group-derived different front, Franks et al. (2009) present a method to
data. Most of the examples are for mammals; this reflects, generate user-controlled ensembles of random networks that
as much as anything, that there are a good number of long- they hope will form the foundation of a framework to
term studies of mammals that contain information ripe for develop a quantitative network sampling methodology.
network studies. The toy network in Fig. 3 also illustrates other structural
One such example in the volume is the analysis of the features which may be explored via network analysis, and
association patterns in a long-term study of a group of which may be of great biological importance. Nodes a–
spider monkeys (Ateles geoffroyi) (Ramos-Fernández et al. h and i–l form clusters of nodes (communities in network
2009). The authors used lagged association rates parlance) more densely connected among themselves than
(Whitehead 2008), a classic “association data” measure, to to others. Many methods for detecting such communities
look at the temporal stability (over years) of relationships have been developed both in the networks and the animal
among females plus the more obviously network-derived association literature (see Croft et al. 2008), and these too
measures of node strength and eigenvector centrality to have been used in the animal sciences to find layers of
differentiate individual social roles. They found that adult social structure in the largely unexplored scale between the
females as a class are at the core of the social structure, but group and the population. For example, association matrix
that within that class there is little evidence for anything methods were used by Vonhof et al. (2004) and Wittemyer
other than random associations. et al. (2005) to find multiple levels of social structure in
In a similar vein, Henzi et al. (2009) used a two-pronged bats, Thyroptera tricolor, and elephants, Loxodonta africa-
approach to unravel details of the temporal variation in na, respectively, while Wolf et al. (2007) used an explicitly
relationships among female chacma baboons (Papio ham- network-based approach to find unexpected layers of social
adryas ursinus). The authors measured lagged association structure in a colony of sea lions, Zalophus wollebaeki.
rates and fed these into eight alternate models of social In this volume, Lusseau and Conradt (2009) made use of
structure. In addition, they constructed weighted association a community analysis as part of their study of bottlenose
networks and used individual node measures to quantify dolphins (Tursiops sp.). Social animals can benefit from
social roles. Each approach led to the conclusion that there group living by combining diverse information to make
is seasonal variation in the extent to which female baboons consensus decisions (Conradt and Roper 2005). This area
maintain differentiated relationships, results which chal- has seldom been explored in the context of social networks.
lenge established views about primate societies. Lusseau and Conradt (2009) looked for evidence of
Also in this volume, Fischhoff et al. (2009) looked at “unshared consensus decisions” among bottlenose dol-
associations among adult male plains zebra (Equus burch- phins. They brought together behavioural observations
elli) over a 4-year period. They are interested in the associated with a collective switch in patch searching with
question of whether the reproductive status of the males network analysis of which animals form clusters or
(bachelor or stallion) has an effect on social groupings and communities, and which are likely to be best informed,
bond formation. In order to test this, the authors developed plus a model of consensus decision making to deduce that
a permutation test that compares bachelors and stallions well-informed individuals can induce a collective decision
while controlling for temporal patterns in grouping and by the group in this species.
Behav Ecol Sociobiol (2009) 63:967–973 971

One of the features of this last example is its use of a Vertebrates Invertebrates
quantitative model to try to tease out what is happening in
PATTERN PROCESS
the data. There are many examples of the use of null
models in network analysis, and rightly so, as these are
often necessary to show that a relational data set contains PROCESS PATTERN
something of biological interest. Examples of models trying
to do more are rather rarer. One nice example is the work of Fig. 4 Schematic representation of the current emphasis in network
Ramos-Fernández et al. (2006) which used an individual- analyses applied to vertebrate and invertebrate social systems
based model of foraging in an environment where feeding
patches vary in size. This simple model produces richly we are more or less forced to observe associations or
structured networks even when there is nothing but simple interactions which, when amalgamated, yield patterns that
aggregation bringing animals together. There is plenty of hopefully make biological sense. Having found these
room in the future for more models of network formation patterns, we naturally ask what effect they will have on,
and dynamics. among other things, process. Assuming that the patterns we
It has long been realised that we can use our understanding find are representative, we ask what effect they will have on
of social patterns to look at the implication for processes such the transmission of information or disease through a
as information transfer or disease transmission (Corner et al. population, for example. By contrast, network studies of
2003; Cross et al. 2004). In this volume, Godfrey et al. invertebrates, of which there are now several (Fewell 2003;
(2009) made a more direct attempt to map disease pathways Naug 2008, 2009), tend to proceed the other way round. In
between individual gidgee skinks (Egernia stokesii). Their most cases, it is (an instance of) the process, such as the
interactions are again associative, via shared use of rock passage of food through a colony, that is observed directly.
crevices. However, the real interest of the authors is in The patterns exhibited by the process tend to come second.
parasite transmission, which they addressed by constructing It is also worth noting that the networks are not group
a “transmission network”, in which two lizards are connected derived, but none of these differences prevents the use of
if they used the same crevice within an estimated transmis- network analysis as the tool of choice.
sion time for parasite infection. This novel approach allowed There is one example of this type of research in this
the authors to begin to analyse the interplay between parasite volume. Naug (2009) presented networks of mouth-to-
load and social position in this species. mouth contacts in nine colonies of the social wasp
Of course, the network paradigm will become much Ropalidia marginata. He found that most colony members
more generally appealing if it can be used for making have a similar number of contacts, but there are relatively
predictions, and there are a couple of convincing studies few contacts between behavioural classes. He then manip-
where this has been achieved. Flack et al. (2006) predicted ulated the colonies to test the resilience of the interaction
the effect of social policing on the structure of interactions networks to removal of individuals. Manipulation and
among a group of primates, and McDonald (2007) showed replication of this kind is rare in network studies but a
that the network connectivity of young male long-tailed particularly promising area for further research.
manakins (Chiroxiphia linearis) predicted their future The examples used in this volume do not begin to
breeding success. In this volume, McDonald (2009) exhaust all the possible uses of network analysis in the
followed up on this earlier study by asking whether it is behavioural sciences. For example, we have not touched
kinship that shapes the “young-boy” network of the earlier upon the possible contributions of network analyses to the
work. The answer, as the author points out, is a rather comparative method, though this is a potentially rich vein.
intriguing “no”. Examples so far in this area include the work by Croft et al.
Another example of network prediction, albeit on a (2006) on guppies and the comparison of two closely
much shorter timescale, appears in the contribution to this related ungulate species by Sundaresan et al. (2007). Much
volume by Eagle and Pentland (2009). These authors of the innovation in network analysis and its application
looked at the underlying structure in daily patterns of have so far come from other disciplines such as sociology
human behaviour, using models. The structure was repre- and statistical physics (Wasserman and Faust 1994;
sented through what the authors call “eigenbehaviors”, a set Newman 2003; see Croft et al. 2008 for a review). In
of characteristic vectors within the multimodal data set addition to novel statistical descriptors for networks, there
detailing the daily routines of individuals and groups. are many new ideas, methods and testable hypotheses to be
In parallel with the development of network methods to found particularly in the sociology literature. For instance,
look at the structure and development of vertebrate Granovetter’s (1973) idea of the importance of weak links
societies, those working with invertebrates have taken a (for information transmission) has yet to be tested in animal
slightly different path (Fig. 4). When studying vertebrates, systems.
972 Behav Ecol Sociobiol (2009) 63:967–973

Of course, we behavioural scientists are not without our Franks DW, Ruxton GD, James R, Noble J (2009) Developing a
methodology for social network sampling. Behav Ecol Sociobiol.
own ideas and methods. Readers are referred to books by
doi:10.1007/s00265-009-0729-2
Whitehead (2008) and Croft et al. (2008), plus a number of Godfrey SS, Bull CM, James R, Murray K (2009) Network structure
recent reviews, including those by Krause et al. (2007) and and parasite transmission in a group-living lizard, the gidgee
Wey et al. (2008). This volume includes a view of the skink, Egernia stokesii. Behav Ecol Sociobiol. doi:10.1007/
s00265-009-0730-9
future possible uses of network theory for behavioural
Granovetter M (1973) The strength of weak ties. Am J Sociol
ecologists (Sih et al. 2009) which, we hope, will whet the 78:1360–1380
appetite of any readers wondering whether network analysis Henzi SP, Lusseau D, Weingrill T, van Schaik CP, Barrett L (2009)
might be of use to them. Cyclicity in the structure of female baboon social networks.
Behav Ecol Sociobiol. doi:10.1007/s00265-009-0720-y
Hinde RA (1976) Interactions, relationships and social structure. Man
11:1–17
James R, Croft DP, Krause J (2009) Potential banana skins in animal
Acknowledgements We would like to thank everyone who took part social network analysis. Behav Ecol Sociobiol. doi:10.1007/
in the IEC symposium on Animal Social Networks in Halifax, Nova s00265-009-0742-5
Scotia, August 2007, as well as Darren Croft, David McDonald and Krause J, Ruxton GD (2002) Living in groups. Oxford University
Dhruba Naug for stimulating discussions. JK acknowledges financial Press, Oxford
support from the NERC (NE/D011035/1) and the EPSRC (GR/ Krause J, Croft DP, James R (2007) Social network theory in the
T11241/01(P)). behavioural sciences: potential applications. Behav Ecol Socio-
biol 62:15–27
Krause S, Mattner L, James R, Guttridge T, Corcoran MJ, Gruber SH,
Krause J (2009) Social network analysis and valid Markov Chain
Monte Carlo tests of null models. Behav Ecol Sociobiol.
doi:10.1007/s00265-009-0746-1
Lusseau D (2003) The emergent properties of a dolphin social
References network. Proc Roy Soc Lond B 270:186–188
Lusseau D, Conradt L (2009) The emergence of unshared consensus
Biggs N, Lloyd E, Wilson R (1986) Graph Theory, 1736–1936. decisions in bottlenose dolphins. Behav Ecol Sociobiol.
Oxford University Press, Oxford doi:10.1007/s00265-009-0740-7
Borgatti SP, Everett MG, Freeman LC (2002) UCINET for Windows, Lusseau D, Whitehead H, Gero S (2008) Applying network methods to
version 6: software for social network analysis. Analytic the study of animal social structures. Anim Behav 75:1809–1815
Technologies, Harvard Maynard Smith J (1982) Evolution and the theory of games.
Camazine S, Deneubourg J, Franks NR, Sneyd J, Theraulaz G, Cambridge University Press, Cambridge
Bonabeau E (2001) Self-organization in biological systems. McDonald DB (2007) Predicting fate from early connectivity in a
Princeton University Press, Princeton social network. PNAS 104:10910–10914
Conradt L, Roper TJ (2005) Consensus decision making in animals. McDonald DB (2009) Young-boy networks without kin clusters in a
Trends Ecol Evol 20:449–456 lek-mating manakin. Behav Ecol Sociobiol doi:10.1007/s00265-
Corner LAL, Pfeiffer DU, Morris RS (2003) Social-network analysis 009-0722-9
of Mycobacterium bovis transmission among captive brushtail Naug D (2008) Structure of the social network and its influence on
possums (Trichosurus vulpecula). Prev Vet Med 59:147–167 transmission dynamics in a honeybee colony. Behav Ecol
Couzin ID, Krause J (2003) Self-organisation and collective behaviour Sociobiol 62:1719–1725
of vertebrates. Adv Study Behav 32:1–67 Naug D (2009) Structure and resilience of the social network in an
Croft DP, Krause J, James R (2004) Social networks in the guppy insect colony as a function of colony size. Behav Ecol Sociobiol
(Poecilia reticulata). Proc Roy Soc Lond B 271:516–519 doi:10.1007/s00265-009-0721-x
Croft DP, James R, Thomas P, Hathaway C, Mawdsley D, Laland KN, Newman MEJ (2003) The structure and function of complex
Krause J (2006) Social structure and co-operative interactions in networks. SIAM Rev 45:167–256
a wild population of guppies (Poecilia reticulata). Behav Ecol Ohtsuki H, Hauert C, Lieberman E, Nowak MA (2006) A simple rule
Sociobiol 59:644–650 for the evolution of cooperation on graphs and social networks.
Croft DP, James R, Krause J (2008) Exploring animal social networks. Nature 441:502–505
Princeton University Press, Princeton Ramos-Fernández G, Boyer D, Gómez VP (2006) A complex social
Cross PC, Lloyd-Smith JO, Bowers JA, Hay CT, Hofmeyr M, Getz structure with fission–fusion properties can emerge from a simple
WM (2004) Integrating association data and disease dynamics in foraging model. Behav Ecol Sociobiol 60:536–549
a social ungulate: bovine tuberculosis in African buffalo in the Ramos-Fernández G, Boyer D, Aureli F, Vick LG (2009) Association
Kruger National Park. Ann Zool Fenn 41:879–892 networks in spider monkeys (Ateles geoffroyi). Behav Ecol
Eagle N, Pentland A (2009) Eigenbehaviors: identifying structure in Sociobiol. doi:10.1007/s00265-009-0719-4
routine. Behav Ecol Sociobiol. doi:10.1007/s00265-009-0739-0 Santos FC, Santos MD, Pacheco JM (2008) Social diversity promotes
Fewell JH (2003) Social insect networks. Science 301:1867–1870 the emergence of cooperation in public goods games. Nature
Fischhoff IR, Dushoff J, Sundaresen SR, Cordingly JE, Rubenstein DI 454:213–216
(2009) Reproductive status influences group size and persistence Scott J (2000) Social network analysis. Sage, London
of bonds in male plains zebra (Equus burchelli). Behav Ecol Sih A, Hanser SF, McHugh KA (2009) Social network theory: new
Sociobiol. doi:10.1007/s00265-009-0723-8 insights and issues for behavioural ecologists. Behav Ecol
Flack JC, Girvan M, de Waal FBM, Krakauer DC (2006) Policing Sociobiol. doi:10.1007/s00265-009-0725-6
stabilizes construction of social niches in primates. Nature Sundaresan SR, Fischhoff IR, Dushoff J, Rubenstein DI (2007)
439:426–429 Network metrics reveal differences in social organization
Behav Ecol Sociobiol (2009) 63:967–973 973

between two fission–fusion species, Grevy's zebra and onager. Whitehead H, Dufault S (1999) Techniques for analyzing vertebrate
Oecologia 151:140–149 social structure using identified individuals: review and recom-
Vonhof MJ, Whitehead H, Fenton MB (2004) Analysis of Spix’s disc- mendations. Adv Study Behav 28:33–74
winged bat association patterns and roosting home ranges reveal Wilson EO (1975) Sociobiology: the new synthesis. Harvard
a novel social structure among bats. Anim Behav 68:507–521 University Press, Cambridge
Wasserman S, Faust K (1994) Social network analysis: methods and Wittemyer G, Douglas-Hamilton I, Getz WM (2005) The socio-
applications. Cambridge University Press, Cambridge ecology of elephants: analysis of the processes creating multi-
Wey T, Blumstein DT, Shen W, Jordán F (2008) Social network tiered social structures. Anim Behav 69:1357–1371
analysis of animal behaviour: a promising tool for the study of Wolf JBW, Mawdsley D, Trillmich F, James R (2007) Social structure
sociality. Anim Behav 75:333–344 in a colonial mammal: unravelling hidden structural layers and
Whitehead H (2008) Analyzing animal societies: quantitative methods their foundations by network analysis. Anim Behav 74:1293–
for vertebrate social analysis. University of Chicago Press, Chicago 1302