Network modularity promotes cooperation

Download Network modularity promotes cooperation

Post on 08-Dec-2016

213 views

Category:

Documents

1 download

Embed Size (px)

TRANSCRIPT

  • d h

    div

    es p

    vidu

    net

    en

    evol

    ork

    ring

    ratio

    ng g

    pr

    Axelrod and Hamilton, 1981). In the prisplayers have a choice between cooperatioplayer receives a pay-off depending on hisof the other player in the game. The highesachieved by defecting regardless of theplayer (Rapoport and Chammah, 1965; Trthe total pay-off for the two players is the h

    , 1965ases inpopula

    conteir pop

    in structured populations in which their associations are not

    that cooperation can evolve in structured populations when

    (Sueur et al., 2011; Whitehead and Lusseau, 2012). This measure

    should favour the evolution of cooperation (Voelkl and Kasper,

    Contents lists available at SciVerse ScienceDirect

    .e

    or

    Journal of Theoretical Biology 324 (2013) 103108with others by controlling the frequency distribution of the populations structure, the two concepts are different. Measures ofcooperators interact more frequently with each other than withdefectors, and share the benets of mutual cooperation (Rapoportand Chammah, 1965; Trivers, 1971). Some games were simulatedon networks in which individuals could interact at different rates

    2009) because cooperation in animals tends to occur amongindividuals of the same social unit (Clutton-Brock et al., 2001;Awata et al., 2010).

    Even though network modularity can provide a measure of

    0022-5193/$ - see front matter & 2012 Elsevier Ltd. All rights reserved.

    http://dx.doi.org/10.1016/j.jtbi.2012.12.012random (Underwood, 1981; Pepper et al., 1999; Newman, 2001;Lusseau, 2003; Croft et al., 2004). Computer simulations ofcooperation games on lattices or networks have demonstrated

    captures a key feature of social networks composed of differentsocial units or communities (Girvan and Newman, 2002; Pallaet al., 2005; Lusseau et al., 2006). Networks with high modularity(Nowak and May, 1992; Ohtsuki et al., 2006). While in thetraditional game players are equally likely to meet with otherplayers (Axelrod and Hamilton, 1981), animals and humans live

    topology that is not inuenced by previously simulated networkfeatures (Reichardt and Bornholdt, 2007; Cao et al., 2011) butplays a key role in enabling the social behaviours of individualscooperate (Rapoport and Chammahcost of the cooperative action increthe percentage of cooperators in aand Doebeli, 2004).

    Cooperation can evolve in a gameinteract randomly, i.e. when theh, 1965; Trivers, 1971;oners dilemma game,n and defection. Eachchoice and the choicet pay-off for a player isdecision of the otherivers, 1971). However,ighest when they both; Trivers, 1971). As therelation to its benet,tion decreases (Hauert

    xt when players do notulation is structured

    and can be estimated using a modularity coefcient (Q) rangingfrom 0 to 1 (Appendix A). A Q close to 1 indicates a network witha strong clustered structure in which interactions of individualsbelonging to different clusters do not occur (Newman, 2006).Modularity can emerge without complicated rules of interactions(e.g. network motif, Milo et al., 2002; hierarchical organisation,Barabasi et al., 2003) but simply from network nodes (individualsin our case) living in varying environments. For example, foodavailability or predator presence might affect individual interac-tions (Stanford, 1995; Heithaus and Dill, 2002). In addition,individuals characteristics such as sex, age, and conditions canshape the way individuals interact with each other (Berman,1982; McPherson et al., 2001; Ruckstuhl, 2007; Marcoux et al.,2010). Modularity is a key characteristic of social networkhas been repeatedly used to model cooperation in populations ofselsh individuals (Rapoport and Chammainto clusters and the degree with which those clusters interact,Letter to Editor

    Network modularity promotes cooperation

    a r t i c l e i n f o

    Keywords:

    Evolution

    Game theory

    Social network

    a b s t r a c t

    Cooperation in animals an

    the evolution of selsh in

    evolve when the game tak

    interactions between indi

    characteristic of all social

    cooperation has never be

    modularity promotes the

    ing games on social netw

    between individuals favou

    for the evolution of coope

    or punishment, or requiri

    wider social contexts than

    1. Introduction

    Cooperation in animals and human is widely observed (Hill,2002; Clutton-Brock, 2009; Dufour et al., 2009; Awata et al., 2010)even if evolutionary biology theories predict the evolution ofselsh individuals (Darwin, 1859). The prisoners dilemma game

    journal homepage: www

    Journal of Theumans is widely observed even if evolutionary biology theories predict

    iduals. Previous game theory models have shown that cooperation can

    lace in a structured population such as a social network because it limits

    als. Modularity, the natural division of a network into groups, is a key

    works but the inuence of this crucial social feature on the evolution of

    investigated. Here, we provide novel pieces of evidence that network

    ution of cooperation in 2-person prisoners dilemma games. By simulat-

    s of different structures, we show that modularity shapes interactions

    the evolution of cooperation. Modularity provides a simple mechanism

    n without having to invoke complicated mechanisms such as reputation

    enetic similarity among individuals. Thus, cooperation can evolve over

    eviously reported.

    & 2012 Elsevier Ltd. All rights reserved.

    number of partners individuals had (Wu et al., 2010; Cao et al.,2011).

    Modularity is a simple feature of all biological networks thatinuences heterogeneity in contacts between nodes in a network(Ravasz et al., 2002; Kashtan and Alon, 2005; Whitehead andLusseau, 2012). Modularity describes the separation of networks

    lsevier.com/locate/yjtbi

    etical Biology

  • population structure can be spatial, genetic or social (e.g. Hinde,1976; Weir and Cockerham, 1984; Slatkin, 1987; Bohonak, 1999).In spatial and genetic structure, interactions between individualsare often modelled homogeneously, i.e. individuals interact withall individuals within their group and do not interact withindividuals outside their group, or if they do, the rate of interac-tion is assumed to be equal among all groups (Killingback et al.,2006; Vainstein et al., 2007; Smaldino and Schank, 2012). Whilemodularity might also arise from spatial or genetic structure, itallows us to model population structure more realistically.Modularity is a continuous measure; interactions between indi-viduals vary continuously; individuals do not exclusively interactwith individuals within their group and do not have to interactwith all individuals in their group. In addition, the rate ofinteraction between groups is not the same for all groups(Newman and Girvan, 2004; Newman, 2006). Variation in therates of association between members of a population, measuredby association indices, also create structure in a population(Hinde, 1976; Whitehead, 1995; Bejder et al., 1998). Associationindices can be binary or continuous, and are used to buildsociograms and social networks (Lusseau, 2003; Whitehead,1999; Croft et al., 2011). Different metrics are computed from

    the association indices to describe the social structure of apopulation (e.g. Whitehead and Dufault, 1999) and modularityis one of them.

    2. Material and methods

    Given the importance of modularity in social networks, weexplored how it inuences the evolution of cooperation in the2-person prisoners dilemma game. We simulated games onweighted networks with varying modularity (Whitehead andLusseau, 2012) in which vertices represented players and edgesrepresented the associations between them (Fig. 1). For mostanalyses, network size was set to 100 players but the sensitivityof outcomes to network sizes was also evaluated by simulatinggames on network of size ranging from 20 to 500 (for details, seeAppendix A). In addition, the effects of cluster size (from 2 to 33)and cost-to-benet ratios (from 0.1 to 0.9) were examined(Appendix A). Games started with an equal number of coopera-tors and defectors, randomly located in the network. A playerentered a game with all its neighbours that had an associationhigher than a threshold value set at the beginning of each

    y blu

    ion

    ma

    lowe

    Letter to Editor / Journal of Theoretical Biology 324 (2013) 103108104Fig. 1. Examples of weighted networks of 50 players (cooperators are represented b(b) and (d) high modularity of 0.62. Networks (a) and (b) show the initial distribut

    cooperators and defectors at the end of 6000 rounds of 2-person prisoners dilem

    modularity conditions of (b) and (d) limited interactions between players, which alcolor in this gure legend, the reader is referred to the web version of this article.)e circles and defectors by red squares) with (a) and (c) low modularity of 0.12; and

    of cooperators and defectors in the networks; (c) and (d) show the distribution of

    games played with a threshold of 0.4 and a cost-to-benet ratio of 0.2. The high

    d for the evolution clusters of cooperators. (For interpretation of the references to

  • simulation. We used threshold values ranging from 0.1 to 0.9(Appendix A) to make our simulations comparable with pre-viously published studies in which the association betweenindividuals was binary (not weighted: Ohtsuki et al., 2006;Nowak et al., 2010; Allen et al., 2012). The evolutionary dynamicsfollowed the deathbirth update rule (Ohtsuki et al., 2006). Weran each simulation for 6000 rounds because this number ofrounds was sufcient to reach equilibrium (Appendix B). Thepercentage of cooperators was averaged for the last 1000 simula-tions. Results s