Behavioral The official journal of the

Ecology ISBE
International Society for Behavioral Ecology

Behavioral Ecology (2014), 25(3), 435–442. doi:10.1093/beheco/aru014

Invited Anniversary Essay

Complexity and behavioral ecology

Jack W. Bradbury and Sandra L. Vehrencamp
Department of Neurobiology and Behavior, Cornell University, Ithaca, NY 14853, USA

Downloaded from http://beheco.oxfordjournals.org/ at Northern Arizona University on December 12, 2014

Received 15 January 2014; accepted 16 January 2014; Advance Access publication 17 February 2014.

Reductionism, the practice of predicting large system properties by adding up the measured behaviors of components, has had a long
and successful run in recent science. However, in the last 2 decades, fields as diverse as physics, ecology, neurobiology, and econom-
ics have recognized that many complicated systems have emergent and self-organized properties that cannot be explained as the
linear sum of components, but instead must be viewed as the potentially diverse outcomes of system nonlinearity. Despite some pio-
neering efforts to apply complexity theory to group movements and decision making, most behavioral ecologists have avoided invoking
complexity perspectives in their research. In this essay, we argue that the reductionist focus on dyads in our field has largely run its
course and that the next frontier is to examine whether and how social and communication networks function as complex nonlinear
systems. To this end, we provide a sampling of topics within behavioral ecology where we think complexity theory may be illuminating.
Key words:   adaptive dynamics, communication, complexity theory, fractals, leks, networks, nonlinear systems, self-organiza-
tion, social markets.

Introduction When the 4 were reunited, the second and third crickets switched
In the early days of ethology and animal behavior research, domi- ranks. These results suggested that crickets used recent contest suc-
nance hierarchies were “the rage.” Although chickens and pri- cess as a rule of thumb to estimate their relative fighting ability. In
mates received the most attention, dominance hierarchies were the absence of better information, this was not a bad option. But
soon found in a wide variety of taxa. And in a surprising num- it could lead to mistakes: Should 1 of 2 equally competent fight-
ber of species, the hierarchies were structurally linear: The α ani- ers slip during a confrontation and then lose the contest, the loser
mal always dominated everyone, the β dominated everyone but α, would reset its self-rank lower and the winner would adjust its self-
etc. Although many researchers took this linearity of hierarchies rank higher. Such adjustments would “spread out” what were oth-
for granted, some even suggesting it had been favored by selec- erwise overlapping rankings for this pair of animals. The adjusted
tion to minimize conflicts, there was a serious problem underlying self-ranks would then bias each animal’s willingness to escalate in
such linear orderings. Landau (1951a, 1951b) was an early critic subsequent interactions with other group members. Although this
who showed that it was mathematically impossible for animals that might stabilize dyadic relationships, it is not immediately obvious
settled contests based on normally distributed traits such as body whether the cumulative interactions among many group members
size to sort themselves into a linear order. The large numbers of relying on such a process would stabilize and generate the observed
animals at and near the average body size would have no accu- linear hierarchies.
rate way to do so. Although Landau’s concern was dutifully noted In fact, recent models have shown that this process can con-
in later reviews and textbook treatments of the topic, its general verge to a stable hierarchy (Beacham 2003; Chase and Seitz 2011).
significance was rarely discussed. The issue surfaced again when In these models, the final hierarchies are more likely to be linear
Chase (1974, 1985, 1986) made a similar point decades later. But when relationships are first established sequentially within sub-
again, the broader implications were often ignored by subsequent groups, and the functions relating combatant self-ranking to the
researchers. probability of winning a contest are nonlinear. Put more generally,
An early clue to resolving this puzzle was provided by Alexander the linearity of hierarchies is here seen as an emergent property,
(1961). He let 4 cricket males establish a linear hierarchy and then an artifact if you will, of complex nonlinear interactions between
isolated the 2 top-ranked animals together and the 2 bottom ani- group members.
mals together. The second-ranked cricket now lost all contests, as Given the early concerns and clues, why has it taken so long for
there was no third-ranking animal for it to dominate. The third- this interpretation of hierarchy linearity to appear? We think the
ranking cricket now won all contests with the lowest ranking one. major impediment was the obsession with reductionist science that
arose in the 1960s and is only now beginning to yield to alternative
approaches. The basic assumption of reductionism is that compli-
Address correspondence to J.W. Bradbury. E-mail: jwb25@cornell.edu. cated systems can be understood by characterizing the properties

© The Author 2014. Published by Oxford University Press on behalf of

the International Society for Behavioral Ecology. All rights reserved. For
permissions, please e-mail: journals.permissions@oup.com
436 Behavioral Ecology

and behaviors of their component parts and then adding these new values for the key variables are drawn at random from some
up to generate the whole. Look at any issue of Behavioral Ecology distribution). The right sides of these equations include extrinsic
or Animal Behaviour and you will find many successful applications parameters that are currently fixed in value. They usually also include
of this approach: Mating systems, parent–offspring conflicts, con- the values of the key variables in the current state. Given some
test behavior, cooperation, and communication are commonly bro- starting values for the key variables, one can successively apply the
ken down into dyadic interactions and the properties of any larger dynamic equation again and again to plot the trajectory of the sys-
ensemble predicted by adding up the dyadic behaviors. tem over time. The trajectory may be quite different depending on
Emergent properties, by definition, do not fit into this paradigm. the initial starting point and the values of the included parameters.
But then, how often does the reductionist strategy fail? Surely, we See Strogatz (1994) for a detailed introduction to dynamic systems
have done pretty well with the reductionist approach for decades. analysis.
Are emergent properties common enough to worry about? In this In a linear system, none of the variables on the right side of the
essay, we argue that it is time for behavioral ecologists to move dynamic equation have any exponents other than 1, there are no
beyond our exclusive focus on dyads and examine how much we products or ratios of key variables, and none of the key variables is
might be missing by not treating complexity head on. Many other present as the argument of a trigonometric, exponential, or similar
fields such as physics, neurobiology, computer science, and econom- function. It is easy to predict the trajectory of a linear system as

Downloaded from http://beheco.oxfordjournals.org/ at Northern Arizona University on December 12, 2014

ics have already made this transition (for a lucid and nonmathe- each key variable changes independently; the next state of the sys-
matical survey, see Strogatz 2004). Thanks to May (1976) and his tem is just the linear sum of the next states of each key variable. If
students, theoretical ecology made this leap long ago. For some rea- a parameter is varied, the system responds proportionally.
son, it has been slow to appear in behavioral ecology. When attend- Reductionism assumes linear systems: Here, you break a system
ing recent ISBE and ABS meetings, we have been surprised at how down into its components, see how each component changes over
few colleagues invoke or even seem aware of complex systems the- time, and add these changes up to predict the overall state of the
ory. There are indeed exceptions, which we note below, but even system at each successive time point. There are no emergent out-
here the researchers seem so focused on their own specific topic comes in a linear system. This does not mean that linear systems
that they often do not make any effort to tie their results into the are boring: Linear system trajectories can progress to an equilib-
general predictions of complexity theory. We do not claim to be rium where further change stops, spiral off into infinity, or exhibit
experts in this approach. But we have learned enough recently to oscillations at some fixed frequency set by the parameters and initial
realize that, were we starting our careers now, this is a key approach conditions. However, each of these trajectories is entirely predict-
in which we would want to invest. We think complexity is one of able given the equations and the values of the extrinsic parameters
the major remaining frontiers in our field. This essay seeks to pro- and initial variable values.
vide some basic background on the topic and suggest just a few of A nonlinear system is one in which one or more of the conditions
the behavioral ecological topics where the application of complex- required for linear systems is violated. Like linear systems, nonlin-
ity approaches may prove enlightening. ear dynamics can move a system to a stable equilibrium point or
spiral off into infinity. However, variation in parameter values may
Some Background not result in proportional variation in the system but, instead, trig-
ger major qualitative changes into totally different states. Such shifts
One problem with any evolving field of research is that workers in state are called bifurcations. Nonlinear systems can exhibit oscilla-
adopting different entry points to that field coin their own names tions, but unlike the harmonic oscillations of linear systems, where
for equivalent or at least overlapping processes. This is certainly the frequency is set by the external parameters and initial condi-
true for complexity theory. Below, we provide an initial definition tions, the oscillations of nonlinear systems are limit cycles whose
of complexity and then summarize several different entry points to frequencies depend on the system itself. If the dimension of the
complexity that are likely to be relevant to behavioral ecology. system is sufficiently large, changing the parameters can cause the
system to go into deterministic or stochastic chaos.
A definition of complexity
A good example of a nonlinear system is the set of vibrating
A number of authors have tried to provide broad definitions of membranes that create signal sounds in vertebrates (Wilden et al.
complexity. For our purposes, we shall adopt a version of one pro- 1998). Consider a terrestrial mammal in which airflow through
posed by Mitchell (2011): Given that a system is an ensemble of the larynx acts as a parameter affecting the paired vocal chords
interacting entities, then a complex system is one that exhibits at least on each side of the flow cavity. At very low flows, the folds remain
some properties that cannot be explained as the linear sum (super- at an immobile equilibrium. At a critical but still moderate flow,
position) of properties of the component elements. The exceptional the thinner parts of each fold begin limit cycle oscillations, sweep-
properties are said to be emergent. This definition sets us up to intro- ing out a repeated 2-dimensional trajectory. Given the moderate
duce our first entry point, nonlinear dynamic systems. flow and their proximity, the 2-folds act as coupled oscillators
and lock into the same frequency. This is the normal vocaliza-
Nonlinear system dynamics tion mode, and because it is periodic but invariably nonsinusoi-
Although behavioral ecologists often focus on stable equilibria (e.g., dal, the resulting sound appears on a spectrogram as a harmonic
ESS analysis), most natural systems in the world can change with series. At a somewhat higher flow, the 2-folds continue to oscil-
time. The temporal changes are called the dynamics of the system. late but the coupling between them breaks down and they may
Typically, one identifies key variables that describe the current state adopt slightly different frequencies. The lower frequency acts as
of the system and then derives equations that predict how these an amplitude modulator of the higher frequency “carrier,” cre-
variables will change in the next time interval. The dimension of the ating sidebands around the higher frequency component in a
system depends on how many key variables are invoked. The equa- spectrogram. This is called biphonation. At even higher flows, the
tions can be deterministic (no chance involved) or stochastic (in which entire complex of vocal folds on each side begins to oscillate, but
Bradbury and Vehrencamp • Complexity and behavioral ecology 437

given the larger masses and the shift to 3-dimensional trajecto- When are animal social or communication networks complex
ries, at a lower frequency. This would be seen on a spectrogram as systems? Given our prior definitions, it should be clear that a net-
a sudden shift from harmonics to subharmonics spaced some frac- work might be either a linear or a nonlinear system: It will depend
tion, often half, of that seen in the prior series. Finally, at high on the nature of the interactive links between network members. If
enough flows, the system lapses into chaos and the spectrogram these relationships are essentially linear, then we would not expect
shows a wide band of noise. One can see several of these modes to see emergent properties and the trajectories followed by these
in 3 successive calls by a wild parrot in Figure 1. Like this parrot, networks should be predictable by knowing the relevant equations,
many animals “bifurcate” their nonlinear vibratory systems from starting points, and ambient parameters. However, there are many
one mode to another when producing signals. Human singers do reasons to believe that the complicated ways that group or network
this on purpose, whereas the “voice break” of teenage boys is an members can affect each other and then be affected in turn by the
unintended bifurcation. resulting feedback will generate nonlinear linkages between indi-
Although the versatility of vertebrate sound-producing organs viduals (see Strogatz quote above). The dominance hierarchies we
is itself interesting to behavioral ecologists studying communica- outlined earlier are a clear case in point. When the network links
tion, the broader message here is that any nonlinear system may are largely nonlinear, then we should not be surprised to see the
be capable of such sudden bifurcations. And nonlinear systems network act like a complex system showing bifurcations and emer-

Downloaded from http://beheco.oxfordjournals.org/ at Northern Arizona University on December 12, 2014

must be common in behavioral ecology. As summarized by Strogatz gent properties like synchronization or other qualitative changes
(1994), “Whenever parts of a system interfere, cooperate, or com- in state. Such emergent behaviors are well known in other kinds
pete, there are nonlinear interactions going on.” So, instead of the of networks such as ecological webs, neurobiological systems, the
structure of sound production, we might see bifurcations such as Internet, and various physical systems (Grossberg 1988; Goldberger
a sudden synchronization of behaviors within a large social group et  al. 1990; Watts and Strogatz 1998; Strogatz 2001; Albert and
(e.g., synchronous firefly flashing), a shift from solitary to aggregated Barabasi 2002; Maslov and Sneppen 2002; Barabási and Bonabeau
dispersions (e.g., migratory locusts), or the rapid breakdown of 2003; Newman 2003; Mitchell 2006; Song et  al. 2006; Gomez-
respect for lek territories leading to chaotic dispersions of display- Gardenes et  al. 2007; Arenas et  al. 2008; Ebeling et  al. 2013;
ing and fighting males (such as we observed with eastern California Suweis et al. 2013).
sage grouse in the acute winter of 1983). Where bifurcations are
possible, they will by definition lead to emergent states, and the Fractals
corresponding nonlinear systems will fit our definition of complex Power laws, in which 1 variable is a function of some other variable
systems. The take-home message is that knowing something about (the argument) raised to some exponent, are common in nature.
nonlinear system dynamics will help us look for possible complexity The inverse square laws for gravity and sound attenuation are
in behavioral ecology. power laws where the exponent is −2. Power laws are self-similar
(also called scale free): A  plot of functional results versus various
Networks values of the argument will have the same shape regardless of the
There has been considerable recent interest in the role of net- scale of values used. Fractals are sets of objects (numbers, points,
work processes in behavioral ecology (see McGregor 2005; Croft lines, etc.) generated by power laws with negative exponents that
et al. 2008; Krause et al. 2009; McDonald 2009; Chapter 15 in are not necessarily integers. The absolute value of the exponent is
Bradbury and Vehrencamp 2011; Pinter-Wollman et  al. 2013). the dimension of the fractal. One can estimate the fractal dimension
Nearly any interacting ensemble of animals can be modeled by plotting the logarithm of the function output against the log of
as a network including primate troops, males on a lek, nesting the argument values; the absolute value of the slope should equal
colonies of seabirds and pinnipeds, communication systems, etc. the fractal dimension.
Because there is already an extensive literature on networks in The set of points defining a straight line on a plane has a dimen-
general (Strogatz 2001; Albert and Barabasi 2002; Newman sion, both classical and fractal, of 1.0. The set of points filling a
2003; Barabási 2009; Pinter-Wollman 2013), measures for clas- bounded area on that plane will have a classical and fractal dimen-
sifying and comparing animal social networks are immediately sion of 2.0. But a squiggling line in the plane that meets fractal
available. criteria will have a fractal dimension between 1.0 and 2.0 because

Figure 1
Nonlinear behaviors in bird vocal organ. Three successive “peow” calls by wild male white-fronted amazon parrot (Amazona albifrons). First call on left shows
typical harmonic series of stable limit cycle vibrations in syrinx. Middle call shows appearance of subharmonics in last third of call (arrow points to relevant
section). Final call lapses into chaos in last two-thirds. Frequency scale (vertical axis): 0–16 kHz; time scale (horizontal axis: 0–1.4 s).
438 Behavioral Ecology

it clearly fills more of the plane than the straight line, but less than Self-organizing systems with few degrees of freedom (e.g., small
the filled area. The power function defining a fractal set can be social networks) might show any of the usual nonlinear outcomes
either deterministic or stochastic; if the latter, the log–log plot will including chaos, whereas larger systems with stronger nonlinearities
show a scattering of points, but we should still be able to discern a may avoid chaos but show self-organizing criticality: Here, the system
fixed exponent from a regression slope. spontaneously evolves to a quasi-stable state of advanced pattern
What do fractals have to do with complexity? It turns out that and complexity but is subject to periodic breakdowns whose mag-
the dynamics of complex systems often lead to fractal sets. The nitudes and/or spatial scales are distributed as power laws and
more structured the fractal set, the lower the fractal dimension; fractals. Thus, small breakdowns in a complex network’s function-
more random sets have higher fractal dimensions. The efficiency of ality might be common, but catastrophic breakdowns would be
functional activities is thought to be optimal for intermediate fractal rare (Bak et al. 1987; Creutz 1992; Bak and Creutz 1994; Turcotte
dimensions. For example, the geometric distributions of blood ves- 1999). The classical physics example is a trickle of dry sand grains
sels and the pulmonary tree have an intermediate fractal dimen- onto a flat surface (Creutz 2004). Over time, a peaked mound of
sion. Biochemical systems with fractal distributions of connections sand builds up on its own: This is a self-organized structure. When
are thought to be more robust to breakdowns than other designs the slope of the pile walls is steep enough, the system becomes self-
(Gallos et al. 2007). Foragers searching for sparse and randomly dis- organized critical. Each subsequent grain of sand added to the pile

Downloaded from http://beheco.oxfordjournals.org/ at Northern Arizona University on December 12, 2014

tributed prey can optimize search by drawing successive step lengths has the potential of triggering an “avalanche.” Most avalanches
from a stochastic fractal distribution with a dimension of about 2.0 will be small, but according to the relevant power law, there is a
(Viswanathan et  al. 2008). Even complex systems that are experi- low but nonzero probability of a major avalanche that flattens the
encing chaos will trace out nonrepeating trajectories that cumula- pile. The notion of self-organized criticality has also been applied
tively obey a fractal rule (Strogatz 1994). For example, the relative to earthquakes, forest fires, disease epidemics, stock market crashes,
proportions of waves of different amplitude in the waveform of a and species diversity collapses in ecological communities.
larynx oscillating chaotically will follow a fractal distribution rule.
An enormous amount of theoretical effort has focused on the Topic overlap
appearance of fractal structures in complex networks (see summary It should be obvious that these various phenomena all show high
in Chapter  15 of Bradbury and Vehrencamp (2011)). Most bio- degrees of overlap. In practice, complexity, self-organization,
logical and human (e.g., Internet) networks show decreasing power and nonlinear dynamics are almost synonymous. Despite this,
functions relating the fraction of nodes in a network to the number researchers focusing on 1 approach often fail to discuss overlap-
of links to a node. However, the log–log plots are not always linear, ping approaches and their relevant literature. Part of the problem
implying that something else is going on. In many cases, for exam- is that several of the key concepts such as complexity, fractals, and
ple, metabolic networks, the network is organized into hierarchical self-organization criticality have fuzzy or highly debated definitions
clusters and modules. Because this same modular structure recurs (Mandelbrot 1983; Turcotte 1999; Mitchell 2011). At the same
at various scales, these networks are structured fractals (low dimen- time, each of these concepts has found important applications
sion). However, other networks include lots of links that reduce the throughout the sciences. We feel that more workers in behavioral
isolation of modules (called “small worlds”; Watts and Strogatz ecology than at present should at least know what they mean and
1998). These tend to have higher fractal dimensions. Current the- how they relate to each other.
ory and data suggest that the more structured networks sacrifice
speed and extent to which local effects are propagated throughout
the network but gain robustness and resilience against functional Possible Applications
breakdowns by isolating key hubs. The less structured systems are Below, we provide just a small sample of research topics in behav-
less functionally resilient, but communication propagates more ioral ecology that we think are particularly likely to benefit (and in
quickly and effectively (Song et  al. 2006; Ay and Krakauer 2007; some cases already are benefitting) from complexity perspectives.
Ay et  al. 2007; Rozenfeld and Makse 2009). The combination Some can be, and even are currently, approximated with linear and
of fractal and network tools thus provides some very interesting dyadic approaches. However, this can only provide part of the pic-
insights to the function and robustness of various network designs. ture and until researchers examine the broader view, we will not
know the degree to which reductionist paradigms are sufficient.
Self-organization
A self-organizing system is one in which “pattern at the global level Communication networks
emerges solely from interactions among lower level components. The earliest work on communication networks largely focused on
Moreover, the rules specifying interactions among the system’s eavesdropping on communicating dyads by third parties (McGregor
components are executed using only local information without 2005). Nonlinear systems with 3 or more variables (i.e., dimensions)
reference to the global pattern. In short, the pattern is an emer- can exhibit a much wider range of trajectories and outcomes than
gent property of the system, rather than a property imposed on systems with only 2 variables (Strogatz 1994). We should thus not
the system by an external ordering influence.” (Camazine et  al. be surprised were we to find that 3 party communication networks
2001). By this and our prior definitions, self-organizing systems are behaved distinctly differently from simple dyads (even if the latter
complex, and models of their dynamics that predict their behavior interact nonlinearly). More recent work has examined the much
well (e.g., that for firefly flash synchronization) are invariably non- larger networks typical of entire social groups (Croft et  al. 2008;
linear. Fractals clearly fit the definition of self-organized systems: Krause et al. 2009). However, the main focus has been the utility of
A fixed power law rule results in highly complex patterns regardless descriptive measures for comparing network structure (Croft et  al.
of scale. Many social networks of animals and people invoke very 2005; Sundaresan et  al. 2007; Hamede et  al. 2009; McDonald
simple and local rules but generate complex global patterns and are 2009; Wolf and Weissing 2010), and not on the potential for
thus self-organized. emergent properties and bifurcations (but see Lusseau 2003). The
Bradbury and Vehrencamp • Complexity and behavioral ecology 439

emergence of synchrony in flashing fireflies and other assemblies that one might observe such an emergent property and then spend
of displaying males is 1 case in which complexity theory has been a lot of effort trying to find some selective and adaptive reason for
applied to animal behavior (Strogatz 2004). Surely there are many its existence when, in fact, it is a direct consequence of being a
more examples of communication networks that are sufficiently nonlinear system. The example of dominance hierarchies should
complex and nonlinear that emergent properties and states may be always be in our minds when faced with unexpected properties and
found once we look for them. traits.

Leks and mating swarms Evolution of cooperation

Leks and mating swarms are subsets of the communication net- It is generally held that the evolution of cooperation is most eas-
works discussed above. However, there is an additional perspective ily achieved in structured populations with limited mixing (Rousset
that may be relevant in their case. As several authors have pointed and Billiard 2000). Some recent work has put a new twist on this
out, many behavioral interactions should be modeled as “markets” idea by viewing a population as a network of interacting individu-
where multiple parties are engaged in negotiations whose outcomes als. The study argues that if the benefit/cost ratio of cooperating is
affect not only a negotiating dyad, but many others as well (Noë greater than the number of network links experienced by the aver-

Downloaded from http://beheco.oxfordjournals.org/ at Northern Arizona University on December 12, 2014

and Hammerstein 1994, 1995; Hammerstein and Hagen 2005). age individual, cooperation can evolve as an emergent property of
Patricelli et  al. (2011) have recently applied such market analo- the network (Ohtsuki et al. 2006; Santos et al. 2006; Nowak et al.
gies to leks. The links between all parties in a market are compli- 2010). Although this claim has stirred up a lot of debate among
cated and likely nonlinear. This may lead to a variety of emergent researchers studying cooperation and altruism (Grafen 2007;
properties. For example, how much of the unanimity of female Taylor et  al. 2007), the basic point remains: higher order proper-
choice on classical leks may be an emergent property similar to the ties such as widespread cooperation or the forming of coalitions are
linearity in dominance hierarchies? Do large leks ever show self- potentially emergent properties given appropriate network structure
organized criticality with power law distributed breakdowns in ter- and nonlinearity. Clearly, there is a lot more than can be mined
ritoriality? Mating swarms of male insects have less structure than from these approaches.
leks, but individuals do adjust their positions according to neighbors
and dominant males often aggregate at the side where females are Personality diversity
most likely to appear (Downes 1969; Thornhill and Alcock 1983; The last decade has seen a burst of interest in the existence and role
Bradbury 1985). Even swarms may constitute sufficiently complex of individual personalities in animals (Sih et al. 2004; Réale et al.
networks that they show emergent properties and bifurcations. We 2007; Dingemanse and Wolf 2010; Réale et al. 2010). One obvious
would not know until we look. question is whether personality diversity is simply a reflection of
residual genetic and phenotypic variation or is instead favored by
Group coordination and decision making selection in some way. Several recent papers suggest a third option
This is 1 area where complexity modeling is already being applied that personality diversity is an emergent property of nonlinear
and proving to be very instructive. The shapes of flying bird flocks interactions between individuals in a population (McNamara et al.
and swimming fish schools are emergent properties of a com- 2009; Botero et  al. 2010). These 2 papers focus on different types
plex system with only local behavior rules (Couzin and Krause of social interactions (trust and communicative accuracy), but they
2003; Hemelrijk and Hildenbrandt 2008; Hemelrijk et  al. 2010; are unlikely to have exhausted the types of interactive networks that
Hildenbrandt et al. 2010). They are thus self-organized. The trails might produce emergent personality diversity. Personality diversity
of ants and the refuges built by social insects are also examples has proved to be so common that there may be other trajectories
of self-organized systems with globally emergent consequences that lead social networks to exhibit this property.
(Camazine et al. 2001). Finally, the processes by which members of
animal groups make joint decisions are often best modeled as com- Foraging dispersions
plex systems with emergent properties (Conradt and Roper 2005; Fretwell and Lucas (1969) defined the “ideal free distribution” as the
Conradt and List 2009; Seeley 2010; Sueur et al. 2011). Although result of settlement on heterogeneously distributed resources when
these examples are breaking new ground, there are clearly many settlers have perfect knowledge, unlimited mobility, and no hin-
other coordination systems in other taxa in which we should be drance to settling except local levels of competition. The final distri-
looking for similar phenomena (or perhaps different outcomes, bution is a Nash equilibrium at which it does not pay for any settled
given system nonlinearity). animal to move elsewhere. This concept has spawned decades of
subsequent theoretical, lab, and field work (see recent reviews in
Colonial breeding Cressman et  al. 2004; Flaxman and deRoos 2007; Křivan et  al.
Many marine vertebrates such as penguins, sea birds, and pinnipeds 2008; Cantrell et al. 2012; van der Hammen et al. 2012; Williams
breed in dense colonies. Terrestrial species such as social weavers, et al. 2013). In the current context, it should be clear that the ideal
bee-eaters, some parrots, some icterids, prairie dogs, hyraxes, etc. free distribution is a self-organized emergent pattern based on local
also breed and sometimes live permanently in colonies. Colonial and individual behaviors. There has been great interest recently in
breeding provides countless opportunities for extrapair mating, which sampling strategies might be optimal for settlers, and which
acquisition of helpers, competition between neighbors, and shared one(s) are actually used in the wild: Truly random walks, corre-
activities such as alarm signaling. All of these interactions can be lated walks, biased random walks, composite walks, Lévy walks,
seen as links in complex networks, and given the opportunities for etc. (Bartumeus and Levin 2008; Raposo et al. 2009; Smouse et al.
feedbacks, many links will be nonlinear. Any of these contexts thus 2010; Viswanathan et  al. 2011; MacIntosh et  al. 2013; Reynolds
has the potential to show emergent properties, state bifurcations, 2013). Interestingly, some models that best fit the data rely on sto-
and other trajectories associated with nonlinear systems. The risk is chastic power function and/or fractal distributions to select either
440 Behavioral Ecology

the distance to move next or how long to wait before changing Arenas A, Diaz-Guilera A, Kurths J, Moreno Y, Zhou CS. 2008.
direction. This is an area where at both the trajectory and final Synchronization in complex networks. Phys Rep. 469:93–153.
Ay N, Flack J, Krakauer DC. 2007. Robustness and complexity co-con-
distribution scales, complexity approaches are likely to provide key structed in multimodal signalling networks. Philos Trans R Soc Lond B
insights. Biol Sci. 362:441–447.
Ay N, Krakauer DC. 2007. Geometric robustness theory and biological net-
Adaptive dynamics works. Theory Biosci. 125:93–121.
Bak P, Creutz M. 1994. Fractals and self-organized criticality. In: Bunde A,
Although ESS thinking now pervades much of behavioral ecology, Havlin S, editors. Fractals in science. New York (NY): Springer-Verlag. p.
it is important to remember that arriving at a stable equilibrium, 26–47.
like an ESS, is only one possible outcome of a system’s evolutionary Bak P, Tang C, Wiesenfeld K. 1987. Self-organized criticality: an explana-
trajectory. Nearly all evolutionary games involve nonlinear dynam- tion of the 1/f noise. Phys Rev Lett. 59:381–384.
Barabási AL. 2009. Scale-free networks: a decade and beyond. Science.
ics, and when more than 2 parties are involved (e.g., more than 2
325:412–413.
dimensions), the possible trajectories and outcomes are diverse. Barabási AL, Bonabeau E. 2003. Scale-free networks. Sci Am. 288:60–69.
A  number of theoreticians have recognized these facts and now Bartumeus F, Levin SA. 2008. Fractal reorientation clocks: linking ani-
study adaptive dynamics in which entire trajectories are traced out mal behavior to statistical patterns of search. Proc Natl Acad Sci USA.
105:19072–19077.

Downloaded from http://beheco.oxfordjournals.org/ at Northern Arizona University on December 12, 2014

based on the relevant dynamic equations and starting conditions.
Beacham JL. 2003. Models of dominance hierarchy formation: effects of
ESSs are not the only fixed (stopping) points in such trajectories, prior experience and intrinsic traits. Behaviour. 140:1275–1303.
and examples of bifurcations, limit cycles, branching trajecto- Botero CA, Pen I, Komdeur J, Weissing FJ. 2010. The evolution of indi-
ries, and chaos can be found depending on the model and initial vidual variation in communication strategies. Evolution. 64:3123–3133.
conditions (Dieckmann and Law 1996; Metz et  al. 1996; Geritz Bradbury JW. 1985. Contrasts between insects and vertebrates in the
et  al. 1997, 1998; Leimar 2005; Nowak 2006; McGill and Brown evolution of male display, female choice, and lek mating. Forts Zool.
31:273–289.
2007; Dercole and Rinaldi 2008; Apaloo and Butler 2009; Leimar Bradbury JW, Vehrencamp SL. 2011. Principles of animal communication.
2009). Good examples of outcomes other than ESSs in evolution- 2nd ed. Sunderland (MA): Sinauer Associates.
ary games are the stable limit cycles involving 3 male morphs in Camazine S, Deneubourg J-L, Franks N, Sneyd J, Theraulaz G, Bonabeau
each of the lizards Uta stansburiana and Lacerta vivipara (Alonzo and E. 2001. Self-organization in biological systems. Princeton (NJ): Princeton
University Press.
Sinervo 2001; Sinervo et  al. 2006, 2007, 2008). A  few other taxa
Cantrell RS, Cosner C, Lou Y. 2012. Evolutionary stability of ideal free dis-
appear to show similar cycles (Frean and Abraham 2001; Kerr persal strategies in patchy environments. J Math Biol. 65:943–965.
et al. 2002), and branching has been suggested as 1 mechanism for Chase ID. 1974. Models of hierarchy formation in animal societies. Behav
generating stable polymorphisms (e.g., personalities) in single popu- Sci. 19:374–382.
lations (Geritz et al. 1998; Leimar 2005). It may be time for more Chase ID. 1985. The sequential-analysis of aggressive acts during hierarchy
formation—an application of the jigsaw puzzle approach. Anim Behav.
of us to step back from ESSs and consider the possibilities of other 33:86–100.
nonlinear evolutionary trajectories. Chase ID. 1986. Explanations of hierarchy structure. Anim Behav.
34:1265–1266.
Chase ID, Seitz K. 2011. Self-structuring properties of dominance hier-
Concluding Remarks archies: a new perspective. In: Huber R, Bannasch DL, Brennan P,
Frishman K, editors. Aggression (Advances in genetics book 75). New
There are certainly additional phenomena within the purview of
York (NY): Academic Press. p. 51–81.
behavioral ecology that are complex systems and may show emer- Conradt L, List C. 2009. Group decisions in humans and animals: a survey.
gent self-organized patterns with all the possible side effects and Philos Trans R Soc Lond B Biol Sci. 364:719–742.
complications. Our goal here was only to provide a few examples Conradt L, Roper TJ. 2005. Consensus decision making in animals. Trends
and not an exhaustive list. If this essay caused you to stop and won- Ecol Evol. 20:449–456.
Couzin ID, Krause J. 2003. Self-organization and collective behavior in ver-
der whether your favorite system is actually nonlinear and complex, tebrates. Adv Stud Behav. 32:1–75.
then we have done our job. It may turn out that many phenomena Cressman R, Krivan V, Garay J. 2004. Ideal free distributions, evolutionary
can be adequately described by linearizing the relevant dynamics games, and population dynamics in multiple-species environments. Am
and focusing on dyads. However, our sense is that there are major Nat. 164:473–489.
discoveries awaiting those who take the broader view, look for Creutz M. 1992. On self-organized criticality. Nucl Phys B-Proc Sup.
26:252–256.
emergent properties, and if they find them, try to understand what Creutz M. 2004. Playing with sandpiles. Physica A. 340:521–526.
kind of complex system they are dealing with. As we noted ear- Croft DP, James R, Krause J, editors. 2008. Exploring animal social net-
lier, the addition of this perspective has had major repercussions in works. Princeton (NJ): Princeton University Press.
other fields of science such as physics, ecology, and economics, and Croft DP, James R, Ward AJ, Botham MS, Mawdsley D, Krause J. 2005.
Assortative interactions and social networks in fish. Oecologia. 143:211–219.
maybe it is time for more of us to give it a burl.
Dercole F, Rinaldi S. 2008. Analysis of evolutionary processes: the adap-
tive dynamics approach and its applications. Princeton (NJ): Princeton
Editor-in-Chief: Leigh Simmons
University Press.
Dieckmann U, Law R. 1996. The dynamical theory of coevolution: a deri-
References vation from stochastic ecological processes. J Math Biol. 34:579–612.
Dingemanse NJ, Wolf M. 2010. Recent models for adaptive person-
Albert R, Barabasi AL. 2002. Statistical mechanics of complex networks. ality differences: a review. Philos Trans R Soc Lond B Biol Sci.
Rev Mod Phys. 74:47–97. 365:3947–3958.
Alexander RD. 1961. Aggressiveness, territoriality, and sexual behavior in Downes JA. 1969. Swarming and mating flight of Diptera. Annu Rev
field crickets (Orthoptera, Gryllidae). Behaviour. 17:130–223. Entomol. 14:271–298.
Alonzo SH, Sinervo B. 2001. Mate choice games, context-dependent good Ebeling W, Feistel R, Feistel S. 2013. On bifurcations in complex ecological
genes, and genetic cycles in the side-blotched lizard, Uta stansburiana. systems with diffusion and noise. Ecol Complex. 14:2–7.
Behav Ecol Sociobiol. 49:176–186. Flaxman SM, deRoos CA. 2007. Different modes of resource variation pro-
Apaloo J, Butler S. 2009. Evolutionary stabilities in multidimensional-traits vide a critical test of ideal free distribution models. Behav Ecol Sociobiol.
and several-species models. Evol Ecol Res. 11:637–650. 61:877–886.
Bradbury and Vehrencamp • Complexity and behavioral ecology 441

Frean M, Abraham ER. 2001. Rock-scissors-paper and the survival of the Stochastic and spatial structures of dynamical systems. Amsterdam
weakest. Proc Biol Sci. 268:1323–1327. (Netherlands): North-Holland Publishing Company. p. 183–231.
Fretwell SD, Lucas HLJ. 1969. On territorial behavior and other factors Mitchell M. 2006. Complex systems: network thinking. Artif Intell.
influencing habitat distribution in birds. Acta Biotheor. 19:16–36. 170:1194–1212.
Gallos LK, Song CM, Makse HA. 2007. A review of fractality and self- Mitchell M. 2011. Complexity: a guited tour. Oxford (UK): Oxford
similarity in complex networks. Physica A. 386:686–691. University Press.
Geritz SAH, Kisdi E, Meszena G, Metz JAJ. 1998. Evolutionarily singu- Newman M. 2003. The structure and function of complex networks. SIAM
lar strategies and the adaptive growth and branching of the evolutionary Rev. 45:167–256.
tree. Evol Ecol. 12:35–57. Noë R, Hammerstein P. 1994. Biological markets: supply and demand
Geritz SAH, Metz JAJ, Kisdi E, Meszena G. 1997. Dynamics of adaptation determine the effect of partner choice in cooperation, mutualism, and
and evolutionary branching. Phys Rev Lett. 78:2024–2027. mating. Behav Ecol Sociobiol. 35:1–11.
Goldberger AL, Rigney DR, West BJ. 1990. Chaos and fractals in human Noë R, Hammerstein P. 1995. Biological markets. Trends Ecol Evol.
physiology. Sci Am. 262:42–49. 10:336–339.
Gómez-Gardeñes J, Moreno Y, Arenas A. 2007. Paths to synchronization on Nowak MA. 2006. Evolutionary dynamics: exploring the equations of life.
complex networks. Phys Rev Lett. 98:034101. Cambridge (MA): Belknap Press.
Grafen A. 2007. An inclusive fitness analysis of altruism on a cyclical net- Nowak MA, Tarnita CE, Wilson EO. 2010. The evolution of eusociality.
work. J Evol Biol. 20:2278–2283. Nature. 466:1057–1062.
Grossberg S. 1988. Nonlinear neural networks: principles, mechanisms, and Ohtsuki H, Hauert C, Lieberman E, Nowak MA. 2006. A simple rule for the

Downloaded from http://beheco.oxfordjournals.org/ at Northern Arizona University on December 12, 2014

architectures. Neural Networks. 1:17–61. evolution of cooperation on graphs and social networks. Nature. 441:502–505.
Hamede RK, Bashford J, McCallum H, Jones M. 2009. Contact networks Patricelli GL, Krakauer AH, McElreath R. 2011. Assets and tactics in a
in a wild Tasmanian devil (Sarcophilus harrisii) population: using social mating market: economic models of negotiation offer insights into animal
network analysis to reveal seasonal variability in social behaviour and its courtship dynamics on the lek. Curr Zool. 57:225–236.
implications for transmission of devil facial tumour disease. Ecol Lett. Pinter-Wollman N, Hobson EA, Smith JE, Edelman AJ, Shiizuka D, de
12:1147–1157. Silva S, Waters JS, Prager SD, Sasaki T, Wittenmyer G, et al. 2013. The
Hammerstein P, Hagen EH. 2005. The second wave of evolutionary eco- dynamics of animal social networks: analytical, conceptual, and theoreti-
nomics in biology. Trends Ecol Evol. 20:604–609. cal advances. Behav Ecol. 25:242–245
Hemelrijk CK, Hildenbrandt H. 2008. Self-organized shape and frontal Raposo EP, Buldyrev SV, da Luz MGE, Viswanathan GM, Stanley
density of fish schools. Ethology. 114:245–254. HE. 2009. Levy flights and random searches. J Physics A.
Hemelrijk CK, Hildenbrandt H, Reinders J, Stamhuis EJ. 2010. Emergence doi:10.1088/1751-8113/42/43/434003.
of oblong school shape: models and empirical data of fish. Ethology. Réale D, Dingemanse NJ, Kazem AJ, Wright J. 2010. Evolutionary and eco-
116:1099–1112. logical approaches to the study of personality. Philos Trans R Soc Lond B
Hildenbrandt H, Carere C, Hemelrijk CK. 2010. Self-organized aerial dis- Biol Sci. 365:3937–3946.
plays of thousands of starlings: a model. Behav Ecol. 21:1349–1359. Réale D, Reader SM, Sol D, McDougall PT, Dingemanse NJ. 2007.
Kerr B, Riley MA, Feldman MW, Bohannan BJ. 2002. Local dispersal Integrating animal temperament within ecology and evolution. Biol Rev
promotes biodiversity in a real-life game of rock-paper-scissors. Nature. Camb Philos Soc. 82:291–318.
418:171–174. Reynolds AM. 2013. Selection pressures give composite correlated random
Krause J, Lusseau D, James R. 2009. Animal social networks: an introduc- walks Lévy walk characteristics. J Theor Biol. 332:117–122.
tion. Behav Ecol Sociobiol. 63:967–973. Rousset F, Billiard S. 2000. A theoretical basis for measures of kin selec-
Krivan V, Cressman R, Schneider C. 2008. The ideal free distribution: a tion in subdivided populations: finite populations and localized dispersal.
review and synthesis of the game-theoretic perspective. Theor Popul Biol. J Evol Biol. 13:814–825.
73:403–425. Rozenfeld HD, Makse HA. 2009. Fractality and the percolation transition
Landau HG. 1951a. On dominance relations and the structure of ani- in complex networks. Chem Eng Sci. 64:4572–4575.
mal societies: I.  effect of inherent characteristics. B Math Biophys. Santos FC, Rodrigues JF, Pacheco JM. 2006. Graph topology plays a deter-
13:1–19. minant role in the evolution of cooperation. Proc Biol Sci. 273:51–55.
Landau HG. 1951b. On dominance relations and the structure of animal Seeley TD. 2010. Honeybee democracy. Princeton (NJ): Princeton
societies: II.  some effects of possible social factors. B Math Biophys. University Press.
13:245–262. Sih A, Bell A, Johnson JC. 2004. Behavioral syndromes: an ecological and
Leimar O. 2005. The evolution of phenotypic polymorphism: randomized evolutionary overview. Trends Ecol Evol. 19:372–378.
strategies versus evolutionary branching. Am Nat. 165:669–681. Sinervo B, Chaine A, Clobert J, Calsbeek R, Hazard L, Lancaster L,
Leimar O. 2009. Multidimensional convergence stability. Evol Ecol Res. McAdam AG, Alonzo S, Corrigan G, Hochberg ME. 2006. Self-
11:191–208. recognition, color signals, and cycles of greenbeard mutualism and altru-
Lusseau D. 2003. The emergent properties of a dolphin social network. ism. Proc Natl Acad Sci USA. 103:7372–7377.
Proc Biol Sci. 270(Suppl 2):S186–S188. Sinervo B, Clobert J, Miles DB, McAdam A, Lancaster LT. 2008. The role
Macintosh AJ, Pelletier L, Chiaradia A, Kato A, Ropert-Coudert Y. 2013. of pleiotropy vs signaller-receiver gene epistasis in life history trade-offs:
Temporal fractals in seabird foraging behaviour: diving through the scales dissecting the genomic architecture of organismal design in social sys-
of time. Sci Rep. 3:1884. tems. Heredity (Edinb). 101:197–211.
Mandelbrot BB. 1983. The fractal geometry of nature. New York (NY): Sinervo B, Heulin B, Surget-Groba Y, Clobert J, Miles DB, Corl A, Chaine
W.H. Freeman and Company. A, Davis A. 2007. Models of density-dependent genic selection and a
Maslov S, Sneppen K. 2002. Specificity and stability in topology of protein new rock-paper-scissors social system. Am Nat. 170:663–680.
networks. Science. 296:910–913. Smouse PE, Focardi S, Moorcroft PR, Kie JG, Forester JD, Morales JM.
May RM. 1976. Simple mathematical models with very complicated 2010. Stochastic modelling of animal movement. Philos Trans R Soc
dynamics. Nature. 261:459–467. Lond B Biol Sci. 365:2201–2211.
McDonald DB. 2009. Young-boy networks without kin clusters in a lek- Song CM, Havlin S, Makse HA. 2006. Origins of fractality in the growth
mating manakin. Behav Ecol Sociobiol. 63:1029–1034. of complex networks. Nat Phys. 2:275–281.
McGill BJ, Brown JS. 2007. Evolutionary game theory and adaptive dynam- Strogatz S. 1994. Nonlinear dyanamics and chaos: with applications to
ics of continuous traits. Annu Rev Ecol Evol Syst. 38:403–435. physics, biology, chemistry, and engineering. Cambridge (MA): Perseus
McGregor PK, editor. 2005. Animal communication networks. Cambridge Book Publishing.
(UK): Cambridge University Press. Strogatz S. 2004. Sync: how order emerges from chaos in the universe,
McNamara JM, Stephens PA, Dall SR, Houston AI. 2009. Evolution of nature, and daily life. New York (NY): Hyperion Books.
trust and trustworthiness: social awareness favours personality differences. Strogatz SH. 2001. Exploring complex networks. Nature. 410:268–276.
Proc Biol Sci. 276:605–613. Sueur C, King AJ, Conradt L, Kerth G, Lusseau D, Mettke-Hofmann C,
Metz JAJ, Geritz SAH, Meszéna G, Jacobs FJA, van Heerwaarden JS. 1996. Schaffner CM, Williams L, Zinner D, Aureli F. 2011. Collective decision-
Adaptive dynamics, a geometrical study of the consequences of nearly making and fission-fusion dynamics: a conceptual framework. Oikos.
faithful reproduction. In: van Strien SJ, Verduyn Lunel SM, editors. 120:1608–1617.
442 Behavioral Ecology

Sundaresan SR, Fischhoff IR, Dushoff J, Rubenstein DI. 2007. Network Viswanathan GM, da Luz MGE, Raposo EP, Stanley HE. 2011. The
metrics reveal differences in social organization between two fission- physics of foraging: an introduction to random searches and biological
fusion species, Grevy’s zebra and onager. Oecologia. 151:140–149. encounters. Cambridge (UK): Cambridge University Press.
Suweis S, Simini F, Banavar JR, Maritan A. 2013. Emergence of structural Viswanathan GM, Raposo EP,da Luz MGE. 2008. Levy flights and super-
and dynamical properties of ecological mutualistic networks. Nature. diffusion in the context of biological encounters and random searches.
500:449–452. Phys Life Rev. 5:133–150.
Taylor PD, Day T, Wild G. 2007. Evolution of cooperation in a finite Watts DJ, Strogatz SH. 1998. Collective dynamics of ‘small-world’ net-
homogeneous graph. Nature. 447:469–472. works. Nature. 393:440–442.
Thornhill R, Alcock J. 1983. The evolution of insect mating systems. Wilden I, Herzel H, Peters G, Tembrock G. 1998. Subharmonics, biphonation,
Cambridge (MA): Harvard University Press. and determinstic chaos in mammal vocalizations. Bioacoustics. 9:171–196.
Turcotte DL. 1999. Self-organized criticality. Rep Prog Phys. 62:1377–1429. Williams AC, Flaherty SE, Flaxman SM. 2013. Quantitative tests of multi-
van der Hammen T, Montserrat M, Sabelis MW, de Roos AM, Janssen A. trophic ideal free distribution theory. Anim Behav. 86:577–586.
2012. Whether ideal free or not, predatory mites distribute so as to maxi- Wolf M, Weissing FJ. 2010. An explanatory framework for adaptive person-
mize reproduction. Oecologia. 169:95–104. ality differences. Philos Trans R Soc Lond B Biol Sci. 365:3959–3968.

Downloaded from http://beheco.oxfordjournals.org/ at Northern Arizona University on December 12, 2014