The role of male contest competition over mates in speciation
The role of male contest competition over mates in speciation
The role of male contest competition over mates in speciation
- No tags were found...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
UNCORRECTED PROOFQVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong><strong>The</strong> <strong>role</strong> <strong>of</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>over</strong> <strong>mates</strong> <strong>in</strong> <strong>speciation</strong>Anna QVARNSTRÖM , Niclas VALLIN, Andreas RUDHAnimal Ecology/ Department <strong>of</strong> Ecology and Genetics, Evolutionary Biology Centre, Uppsala University, Norbyvägen 18D, SE-752 36Uppsala, SwedenAbstract Research on the <strong>role</strong> <strong>of</strong> sexual selection <strong>in</strong> the <strong>speciation</strong> process largely focuses on the diversify<strong>in</strong>g <strong>role</strong> <strong>of</strong> mate choice. Inparticular, much attention has been drawn to the fact that population divergence <strong>in</strong> mate choice and <strong>in</strong> the <strong>male</strong> traits subject to choicedirectly can lead to assortative mat<strong>in</strong>g. However, <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>over</strong> <strong>mates</strong> also constitutes an important mechanism <strong>of</strong> sexualselection. We review recent empirical studies and argue that sexual selection through <strong>male</strong> <strong>contest</strong> <strong>competition</strong> can affect <strong>speciation</strong> <strong>in</strong> waysother than mate choice. For example, biases <strong>in</strong> aggression towards similar competitors can lead to disruptive and negative frequencydependentselection on the traits used <strong>in</strong> <strong>contest</strong> <strong>competition</strong> <strong>in</strong> a similar way as <strong>competition</strong> for other types <strong>of</strong> limited resources. More<strong>over</strong>,<strong>male</strong> <strong>contest</strong> abilities <strong>of</strong>ten trade-<strong>of</strong>f aga<strong>in</strong>st other abilities such as parasite resistance, protection aga<strong>in</strong>st predators and general stresstolerance. Populations experienc<strong>in</strong>g different ecological conditions should therefore quickly diverge non-randomly <strong>in</strong> a number <strong>of</strong> traits<strong>in</strong>clud<strong>in</strong>g <strong>male</strong> <strong>contest</strong> abilities. In resource based breed<strong>in</strong>g systems, a feedback loop between competitive ability and habitat use may lead t<strong>of</strong>urther population divergence. We discuss how population divergence <strong>in</strong> traits used <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> can lead to the build up <strong>of</strong>reproductive isolation through a number <strong>of</strong> different pathways. Our ma<strong>in</strong> conclusion is that the <strong>role</strong> <strong>of</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>in</strong> <strong>speciation</strong>rema<strong>in</strong>s largely scientifically unexplored [Current Zoology 58 ( ): – , 2012].Keywords Male-<strong>male</strong> <strong>competition</strong>, Sexual selection, Speciation, Resource based breed<strong>in</strong>g systems, Contest <strong>competition</strong>, Populationdivergence, Reproductive isolation1 IntroductionA major contemporary goal <strong>in</strong> <strong>speciation</strong> research is to <strong>in</strong>vestigate the mechanisms lead<strong>in</strong>g to population divergenceand reproductive isolation (Coyne and Orr, 2004; Dieckmann et al., 2004; Price, 2008). However, one mechanism <strong>of</strong>selection, i.e. sexual selection through <strong>male</strong> <strong>contest</strong> <strong>competition</strong>, which can lead to fast evolutionary changes, hasreceived surpris<strong>in</strong>gly little attention <strong>in</strong> the context <strong>of</strong> <strong>speciation</strong>. <strong>The</strong> aims <strong>of</strong> this article are to review studies on the <strong>role</strong><strong>of</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>in</strong> <strong>speciation</strong> and to p<strong>in</strong>po<strong>in</strong>t some major unexplored directions for future research.Natural selection leads to the evolution <strong>of</strong> traits that enhance their bearers’ survival chances and reproductive output <strong>in</strong>their natural environment. Sexual selection also leads to the evolution <strong>of</strong> traits enhanc<strong>in</strong>g their bearers’ reproductiveoutput but <strong>in</strong> this case through access to <strong>mates</strong>, which can be viewed as a limit<strong>in</strong>g resource for the sex with the highestpotential reproductive rate (Andersson, 1994). Due to their higher <strong>in</strong>vestment <strong>in</strong> gamete size and therefore <strong>in</strong> <strong>of</strong>fspr<strong>in</strong>gquality (sometimes further <strong>in</strong>creased by pregnancy and maternal care), fe<strong>male</strong>s <strong>of</strong>ten have a lower potentialReceived Nov. 1, 2011; accepted Jan. 20, 2012. Correspond<strong>in</strong>g author. E-mail: anna.qvarnstrom@ebc.uu.se© 2012 Current Zoology1
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>reproductive rate than <strong>male</strong>s. This means that access to <strong>mates</strong> generally does not limit the reproductive rate <strong>of</strong> fe<strong>male</strong>s,while the opposite is true for <strong>male</strong>s (Trivers, 1972). For simplicity we will therefore focus our review on <strong>male</strong> <strong>contest</strong><strong>competition</strong> <strong>over</strong> <strong>mates</strong> rather than on fe<strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>over</strong> <strong>mates</strong>. However, some <strong>of</strong> the matters discussed<strong>in</strong> this article can also apply to fe<strong>male</strong> <strong>contest</strong> <strong>competition</strong> (see e.g. van Doorn et al., 2004)Both natural and sexual selection are known as important mechanisms underly<strong>in</strong>g the process <strong>of</strong> <strong>speciation</strong> butnatural selection has received by far the most scientific attention. Speciation driven by divergent natural selection, <strong>of</strong>tenreferred to as ecological <strong>speciation</strong> (e.g. Schluter, 2000; Rundle and Nosil, 2005), has been identified as the underly<strong>in</strong>gforce driv<strong>in</strong>g whole adaptive radiations whereby a s<strong>in</strong>gle ancestral species splits <strong>in</strong>to a large number <strong>of</strong> new daughterspecies. Adaptive radiations have been documented <strong>in</strong> relation to entry <strong>in</strong>to novel environments, for examples cichlids<strong>in</strong> volcano lakes (Elmer et al., 2010) and f<strong>in</strong>ches exposed to climate change (Grant and Grant, 1993), or <strong>in</strong> relation tothe evolution <strong>of</strong> key <strong>in</strong>novations (i.e. evolution <strong>of</strong> traits that allow the use <strong>of</strong> novel resources, Berenbaum, 1983)whereby the ancestral species splits <strong>in</strong>to daughter species that occupy a large number <strong>of</strong> previous unexplored niches.Strong <strong>competition</strong> <strong>over</strong> limit<strong>in</strong>g natural resources can also by itself lead to splitt<strong>in</strong>g <strong>of</strong> populations. This is because<strong>in</strong>dividuals that are able to utilize resources that are not used by most <strong>in</strong>dividuals <strong>in</strong> the population are favored byselection, i.e. there will be disruptive selection (Dieckmann and Doebeli, 1999).Sexual selection has also been acknowledged as an important evolutionary force driv<strong>in</strong>g <strong>speciation</strong> (West-Eberhard,1983; Price, 1998; Panhuis et al., 2001; Edwards et al., 2005; Ritchie, 2007; Kraaijeveld et al., 2011). <strong>The</strong>re are at leastfour reasons to assume that sexual selection plays an important <strong>role</strong> <strong>in</strong> promot<strong>in</strong>g genetic divergence betweenpopulations and the build up <strong>of</strong> reproductive isolation between them. First, <strong>competition</strong> <strong>over</strong> <strong>mates</strong> can be very <strong>in</strong>tense(Andersson, 1994) and can lead to fast divergence <strong>in</strong> sexually selected traits as compared to naturally selected traits (butsee e.g. Svensson and Gosden, 2007). Second, closely related species <strong>of</strong>ten differ more <strong>in</strong> sexually selected traits than <strong>in</strong>other phenotypic traits, which is expected if these traits are subject to strong selection with<strong>in</strong> species. In addition,divergence <strong>in</strong> sexually selected traits may have been further re<strong>in</strong>forced through selection aga<strong>in</strong>st <strong>in</strong>terbreed<strong>in</strong>g betweenclosely related species (Dobzhansky, 1940; Coyne and Orr, 1989; Howard, 1993; Butl<strong>in</strong>, 1995; Hostert, 1997; Noor,1999; Ortiz-Barrientos et al., 2009). Third, sexually selected traits display extreme variation across animal taxa (West-Eberhard, 1983; Price, 1998; Panhuis et al., 2001; Edwards et al., 2005; Ritchie, 2007; Kraaijeveld et al., 2011)probably as a consequence <strong>of</strong> sexual selection operat<strong>in</strong>g <strong>in</strong> more arbitrary directions than natural selection. Fourth,because fe<strong>male</strong>s <strong>of</strong>ten base their choice <strong>of</strong> mate on sexually selected traits, population divergence <strong>in</strong> such traits maydirectly lead to assortative mat<strong>in</strong>g (e.g. Lande, 1981; Higashi et al., 1999). This latter argument is the ma<strong>in</strong> reason whysexual selection is <strong>of</strong>ten acknowledged as an important component prevent<strong>in</strong>g homogeniz<strong>in</strong>g gene flow betweendiverg<strong>in</strong>g populations when <strong>speciation</strong> occurs <strong>in</strong> sympatry or at secondary contact.Sexual selection through mate choice is unlikely to by itself lead to <strong>speciation</strong> (reviewed <strong>in</strong> Ritchie, 2007). S<strong>in</strong>cesexual selection through mate choice is <strong>of</strong>ten unidirectional <strong>in</strong>stead <strong>of</strong> bidirectional, the <strong>in</strong>itiation and ma<strong>in</strong>tenance <strong>of</strong> adisruptive selection regime is rather unlikely. Many <strong>male</strong> display traits are condition-dependent and signal their bearers’ability to provide their <strong>mates</strong> with good genes or resources needed for reproduction (Andersson, 1994). Environmentspecific costs <strong>of</strong> mate choice may lead to different levels <strong>of</strong> sexual selection and divergence <strong>in</strong> <strong>male</strong> traits <strong>in</strong> allopatricpopulations (Maan and Seehausen, 2011), but would not ensure assortative mat<strong>in</strong>g at secondary contact. For example, iftwo populations diverge <strong>in</strong> <strong>male</strong> body size because fe<strong>male</strong> choice <strong>in</strong> favor <strong>of</strong> large body size acts as a weaker selectionpressure <strong>in</strong> one population, <strong>male</strong>s belong<strong>in</strong>g to the population with large <strong>in</strong>dividuals would nevertheless exert asupernormal sexual stimulus for fe<strong>male</strong>s from the other population at secondary contact (see Labonne and Hendry,2010 for a similar case concern<strong>in</strong>g color <strong>in</strong> guppies). However, Schluter and Price (1993) argued that fe<strong>male</strong> choicecould diverge between allopatric populations when several <strong>male</strong> traits <strong>in</strong>dicate the same abilities but the perception <strong>of</strong>2
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>these different traits varies across environments. Another recent model launched the idea that if only locally adapted<strong>male</strong>s can develop large condition-dependent ornaments fe<strong>male</strong> discrim<strong>in</strong>ation aga<strong>in</strong>st immigrant <strong>male</strong>s would arise(van Doorn et al., 2009). Hence, fe<strong>male</strong> choice based on <strong>in</strong>dicator traits could generate divergence <strong>in</strong> <strong>male</strong> traits orfacilitate assorative mat<strong>in</strong>g under certa<strong>in</strong> conditions.<strong>The</strong> <strong>in</strong>itiation <strong>of</strong> a disruptive selection regime may appear more likely when based on arbitrary Fisherian run-awayprocesses. Given sufficient <strong>in</strong>itial genetic variation <strong>in</strong> fe<strong>male</strong> mate preferences several Fisherian run-away processesmay even operate with<strong>in</strong> the same sympatric population (Higashi et al., 1999). In comb<strong>in</strong>ation with elements <strong>of</strong> fe<strong>male</strong>fe<strong>male</strong><strong>competition</strong> and <strong>male</strong>-<strong>male</strong> <strong>competition</strong>, Fisherian processes could cause disruptive frequency dependentselection (van Doorn et al., 2004). Hence, sympatric <strong>speciation</strong> through Fisherian sexual selection is theoreticallypossible but <strong>in</strong> practice rather unlikely because it requires very specific conditions (van Doorn et al., 2004). When<strong>speciation</strong>, at least partly, occurs with gene flow, there also need to be mechanisms that prevent the break down (due torecomb<strong>in</strong>ation) <strong>of</strong> associations between traits caus<strong>in</strong>g selection aga<strong>in</strong>st hybrids and the traits <strong>in</strong>volved <strong>in</strong> pre-mat<strong>in</strong>gisolation (<strong>in</strong> this case between <strong>male</strong> display traits and fe<strong>male</strong> choice; Kirkpatrick and Ravigné, 2002; Gavrilets, 2004;Servedio et al., 2011). Thus, sexual selection through fe<strong>male</strong> mate choice is recognized as an important evolutionaryforce <strong>in</strong> the context <strong>of</strong> <strong>speciation</strong>, but there are several reasons for assum<strong>in</strong>g that this force leads to <strong>speciation</strong> onlyunder a limited range <strong>of</strong> conditions.Sexual selection is driven not only through mate choice. Male <strong>contest</strong> <strong>competition</strong> <strong>over</strong> access to fe<strong>male</strong>s orresources needed to attract fe<strong>male</strong>s (i.e. breed<strong>in</strong>g territories) is recognized as an important mechanism <strong>of</strong> sexualselection (Andersson, 1994). Male <strong>contest</strong> <strong>competition</strong> favors the evolution <strong>of</strong> a wide range <strong>of</strong> traits <strong>in</strong>clud<strong>in</strong>g weaponssuch as horns and spurs that are directly used <strong>in</strong> combat, sexual size dimorphism, and conspicuous body coloration andcalls that signal fight<strong>in</strong>g ability. Many sexually selected traits have a dual function mean<strong>in</strong>g that they are used both <strong>in</strong><strong>male</strong> <strong>contest</strong> <strong>competition</strong> and for fe<strong>male</strong> mate choice (Berglund et al., 1996; Wong and Candol<strong>in</strong>, 2005). In manymat<strong>in</strong>g systems, <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>over</strong>rides the effect <strong>of</strong> fe<strong>male</strong> choice and these two mechanisms <strong>of</strong> sexualselection can work <strong>in</strong> different or oppos<strong>in</strong>g directions (Qvarnström and Forsgren, 1998; Candol<strong>in</strong>, 2004; Hunt et al.,2009). Intensive <strong>male</strong> <strong>contest</strong> <strong>competition</strong> may even lead to evolution <strong>of</strong> traits that <strong>in</strong>creases <strong>male</strong> mat<strong>in</strong>g success butreduces fe<strong>male</strong> fitness, which <strong>in</strong> turn may trigger an evolutionary arms race between the two sexes whereby fe<strong>male</strong>sevolve counter adaptations (Arnqvist and Rowe, 2002). Thus, <strong>in</strong> a particular population, <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>over</strong>access to <strong>mates</strong> could be: a) the ma<strong>in</strong> mechanism <strong>of</strong> sexual selection, b) a mechanism that sets the stage for which traitsfunction as a target <strong>of</strong> fe<strong>male</strong> choice, or c) a mechanism that triggers a sexually antagonistic co-evolutionary arms-race.Still, the diversify<strong>in</strong>g <strong>role</strong> <strong>of</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> has largely been ignored (but see e.g. Seehausen and Schluter,2004). <strong>The</strong> underly<strong>in</strong>g reason for this ignorance is probably not the assumption that <strong>male</strong> <strong>contest</strong> <strong>competition</strong> is a lesspowerful mechanism <strong>of</strong> sexual selection than fe<strong>male</strong> mate choice. <strong>The</strong> strong focus on fe<strong>male</strong> choice is rather aconsequence <strong>of</strong> its <strong>in</strong>tuitive <strong>role</strong> <strong>in</strong> caus<strong>in</strong>g assortative mat<strong>in</strong>g between diverg<strong>in</strong>g populations. Can <strong>male</strong> <strong>contest</strong><strong>competition</strong> promote <strong>speciation</strong> while fe<strong>male</strong> choice for the same sexually selected character is absent or rema<strong>in</strong>sunchanged? In this review we will argue that the answer to this question is yes.We will focus our discussion on four ma<strong>in</strong> aspects <strong>of</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> and <strong>speciation</strong>:1. Male <strong>contest</strong> <strong>competition</strong> and the evolution <strong>of</strong> novel traits.2. <strong>The</strong> l<strong>in</strong>k between ecology and <strong>male</strong> <strong>contest</strong> <strong>competition</strong>.3. Male <strong>contest</strong> <strong>competition</strong> and character displacement between populations.4. Male <strong>contest</strong> <strong>competition</strong> and reproductive isolation.3
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>We conclude that sexual selection through <strong>male</strong> <strong>contest</strong> <strong>competition</strong> can play an important <strong>role</strong> <strong>in</strong> <strong>speciation</strong>. Basedon this and other conclusions (see below), we would like to encourage more studies focus<strong>in</strong>g on <strong>male</strong> <strong>contest</strong><strong>competition</strong> as a diversify<strong>in</strong>g force and on the <strong>in</strong>teraction between different mechanisms <strong>of</strong> sexual and natural selection<strong>in</strong> the process <strong>of</strong> <strong>speciation</strong> rather than aim<strong>in</strong>g at isolat<strong>in</strong>g their relative importance.2 Negative Frequency-dependent Selection and the Evolution <strong>of</strong> Novel TraitsSpeciation through sexual selection requires that at least one <strong>of</strong> the two splitt<strong>in</strong>g populations evolves a sexuallyselected trait that differs from the ancestral state. Many color patterns are largely affected by a relatively small number<strong>of</strong> genes (Mallet, 1989; Grant and Grant, 1997; Hoekstra et al., 2006). Mutations <strong>in</strong> genes may frequently provide newphenotypic variation <strong>in</strong> coloration and hence potential new targets for sexual selection (Price, 2002; H<strong>of</strong>reiter andSchöneberg, 2010; Manceau et al., 2011). Fe<strong>male</strong> choice may favor such new color patterns due to preexist<strong>in</strong>g biases <strong>in</strong>their sensory systems (Kirkpatrick, 1987), or through learn<strong>in</strong>g (Qvarnström et al., 2004). However, even if sensorybiases or learn<strong>in</strong>g can result <strong>in</strong> fe<strong>male</strong> mate choice <strong>in</strong> favor <strong>of</strong> novel traits, these mechanisms can hardly ma<strong>in</strong>ta<strong>in</strong> stablepolymorphism. Stable trait polymorphism requires negative frequency-dependent selection, i.e. sexual selectionfavor<strong>in</strong>g rare <strong>male</strong> phenotypes. <strong>The</strong>re are some exceptional cases reported <strong>of</strong> fe<strong>male</strong> preferences for rare <strong>male</strong>phenotypes, e.g. <strong>in</strong> guppies (Farr, 1977) but rare phenotypes <strong>of</strong>ten suffer decreased mat<strong>in</strong>g success mean<strong>in</strong>g that matechoice <strong>in</strong>stead generates stabiliz<strong>in</strong>g sexual selection (Kirkpatrick and Nuismer, 2004). By contrast, <strong>male</strong> <strong>contest</strong><strong>competition</strong> has been suggested to more <strong>of</strong>ten fulfill this requirement as <strong>male</strong>s generally bias their aggression towardsrivals that are most similar to themselves s<strong>in</strong>ce such <strong>male</strong>s are most likely to share important resources (van Doorn et al.,2004; Mikami et al., 2004; Seehausen and Schluter, 2004, see Figure 1). This means that <strong>male</strong>s with a rare new traitshould receive less aggression from other <strong>male</strong>s and thereby experience an advantage when establish<strong>in</strong>g breed<strong>in</strong>gterritories. <strong>The</strong>y are also less likely to become <strong>in</strong>jured or to waste their time by fight<strong>in</strong>g with other <strong>male</strong>s <strong>in</strong>stead <strong>of</strong>court<strong>in</strong>g fe<strong>male</strong>s (Hunt<strong>in</strong>gford and Turner, 1987). Thus, a bias <strong>in</strong> <strong>male</strong> aggression towards similar competitors has thepotential to facilitate both the <strong>in</strong>vasion <strong>of</strong> a new trait <strong>in</strong>to the population and to promote stable coexistence <strong>of</strong> differentphenotypes.Pioneer<strong>in</strong>g empirical work on the diversify<strong>in</strong>g <strong>role</strong> <strong>of</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> through negative frequencydependentsexual selection has been done on African cichlids. <strong>The</strong> high diversity <strong>of</strong> haplochrom<strong>in</strong>e cichlids <strong>in</strong> LakeVictoria appears to orig<strong>in</strong>ate from only a few ancestors less than 200 000 years ago (Nagl et al., 2000). Although theseyoung cichlid species have diverged ecologically (Bouton et al., 1997; Seehausen and Bouton, 1997), the fact that thereare strik<strong>in</strong>g differences <strong>in</strong> <strong>male</strong> nuptial coloration has lead to the conclusion that sexual selection has played a central<strong>role</strong> <strong>in</strong> this <strong>speciation</strong> process. Sibl<strong>in</strong>g species with<strong>in</strong> genera, <strong>in</strong> fact, tend to be rather ecologically similar but stronglydifferent <strong>in</strong> <strong>male</strong> nuptial coloration (Seehausen and Schluter, 2004). While sexual selection through fe<strong>male</strong> choice hasbeen given much attention (e.g. Dom<strong>in</strong>ey, 1984; Seehausen and van Alphen, 1999; Knight and Turner, 2004), the <strong>role</strong><strong>of</strong> <strong>male</strong> aggressive <strong>in</strong>teractions for access to <strong>mates</strong> and/or resources necessary to attract <strong>mates</strong> was suggested morerecently. Seehausen and Schluter (2004) exam<strong>in</strong>ed the distribution <strong>of</strong> coloration <strong>in</strong> rock-dwell<strong>in</strong>g haplochrom<strong>in</strong>ecichlids and found that the occurrence <strong>of</strong> species <strong>in</strong> habitat patches was negatively correlated with closely relatedspecies with similar color. Similar patterns <strong>of</strong> non-random distribution <strong>of</strong> <strong>male</strong>s depend<strong>in</strong>g on their color pattern havebeen found <strong>in</strong> Lake Malawi (Young et al., 2009). Dijkstra et al. (2007) carried out aggression tests <strong>in</strong> response to blueand red phenotypes <strong>of</strong> the cichlid species complex Pundamilia (Figure 2). <strong>The</strong> <strong>male</strong> cichlid fish used <strong>in</strong> theseexperiments had been caught from three different types <strong>of</strong> wild populations. Dur<strong>in</strong>g the experimental trials, the wildcaughtfish were placed <strong>in</strong> a test aquarium and allowed to establish a territory. Territorial <strong>in</strong>trusions were then simulatedthrough plac<strong>in</strong>g two transparent watertight tubes conta<strong>in</strong><strong>in</strong>g one red and one blue <strong>in</strong>truder <strong>in</strong> the aquarium together withthe focal fish. <strong>The</strong> outcome <strong>of</strong> these trials differed depend<strong>in</strong>g on where the territorial fish had been caught <strong>in</strong> nature. As4
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>expected, blue <strong>male</strong> Pundamalia cichlids from populations with either a cont<strong>in</strong>uous color morph distribution or with alack <strong>of</strong> <strong>in</strong>termediate forms displayed more aggressive behaviors towards <strong>in</strong>truders belong<strong>in</strong>g to their own color morphthan to the other <strong>in</strong>truder. By contrast, blue <strong>male</strong> cichlids from a population with <strong>in</strong>termediate forms but without acont<strong>in</strong>uous distribution <strong>of</strong> color morphs biased their aggression towards red <strong>in</strong>truders. Why did not blue <strong>male</strong> cichlidsfrom this latter location also bias their aggression towards <strong>male</strong>s belong<strong>in</strong>g to their own color morph? <strong>The</strong> authorssuggested a possible explanation. Red <strong>male</strong> cichlids <strong>in</strong> general behave more aggressively than the blue ones and blue<strong>male</strong> cichlids may therefore have learnt from previous experiences <strong>of</strong> <strong>in</strong>teract<strong>in</strong>g and compet<strong>in</strong>g with such aggressivered <strong>male</strong>s. This explanation is based on the assumption that blue <strong>male</strong>s from the three different populations hadexperienced different levels <strong>of</strong> previous <strong>competition</strong> with red color morphs depend<strong>in</strong>g on where they had been caught <strong>in</strong>the wild. Dijkstra et al (2007) concluded that their results suggest that a new red morph should be able to <strong>in</strong>vade a bluepopulation but that a stable coexistence <strong>of</strong> phenotypes would require additional mechanisms. Hence, social dom<strong>in</strong>anceeffects and learn<strong>in</strong>g seem to play an important <strong>role</strong> <strong>in</strong> haplochrom<strong>in</strong>e cichlid fish (Dijkstra et al., 2007; Dijkstra andGroothuis, 2011) imply<strong>in</strong>g that the frequency-dependent selection caused by <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>of</strong>ten isasymmetric between sibl<strong>in</strong>g species. Thus, two important take home messages are that there may be <strong>in</strong>tr<strong>in</strong>sicdifferences <strong>in</strong> aggressiveness between color morphs and that <strong>in</strong>dividuals may adjust their level <strong>of</strong> aggression towardsparticular color morphs depend<strong>in</strong>g on previous experiences.Experimental work on birds has yielded similar results, i.e. asymmetric color effects on w<strong>in</strong>n<strong>in</strong>g <strong>contest</strong> <strong>competition</strong>.Recently, Pearce et al (2011) <strong>in</strong>vestigated the aggressive response to simulated territory <strong>in</strong>trusions <strong>in</strong> two compet<strong>in</strong>gspecies <strong>of</strong> birds. Male gouldian f<strong>in</strong>ches Erythrura gouldiae were found to be more aggressive towards conspecific<strong>in</strong>truders than towards heterospecific <strong>in</strong>truders and the red head-color morphs were more aggressive than the blackmorphs. By contrast, long tailed f<strong>in</strong>ches Poephilia acuticauda were more aggressive towards gouldian f<strong>in</strong>ches thantowards conspecific models. An <strong>in</strong>tr<strong>in</strong>sic difference <strong>in</strong> aggression between the two species is therefore a likelyexplanation for why long-tailed f<strong>in</strong>ches have an advantage <strong>in</strong> <strong>competition</strong> <strong>over</strong> the limited nest sites needed to attractfe<strong>male</strong>s (Brazil-Boast et al., 2011).<strong>The</strong> l<strong>in</strong>ks between sexually selected traits, <strong>male</strong> aggressive behavior and competitive ability become even moreobvious when the sexually selected trait itself directly affects fight<strong>in</strong>g ability and/or is used as a weapon, e.g. body sizeor horns and spurs. Should we then expect <strong>male</strong> <strong>contest</strong> <strong>competition</strong> to work <strong>in</strong> a unidirectional way with the mostaggressive <strong>male</strong>s always gett<strong>in</strong>g access to most <strong>of</strong> the fe<strong>male</strong>s? Intensive <strong>in</strong>teractions between <strong>male</strong>s <strong>in</strong> densepopulations may cause disruptive selection and favor <strong>male</strong>s with alternative reproductive tactics, e.g. a competitiveversus a sneaker tactic (Figure 1). Such different tactics are generally associated with differences <strong>in</strong> whole suites <strong>of</strong><strong>male</strong> traits associated with fight<strong>in</strong>g ability <strong>in</strong>clud<strong>in</strong>g behavioral, morphological and physiological traits (Gross, 1996).A strik<strong>in</strong>g example is dung beetles that dig burrows underneath animal excrement <strong>in</strong> which they bury portions <strong>of</strong> thedung with eggs attached to them. <strong>The</strong> young will feed on the dung after emerg<strong>in</strong>g from the egg. How much dung theparents have put together for a <strong>male</strong> <strong>of</strong>fspr<strong>in</strong>g will determ<strong>in</strong>e whether he will adopt a competitive or a sneak<strong>in</strong>g tacticwhen reach<strong>in</strong>g sexual maturity. This is because the size <strong>of</strong> the dung ball determ<strong>in</strong>es the adult size <strong>of</strong> the beetle and only<strong>male</strong>s that have reached a certa<strong>in</strong> threshold size will develop horns (Moczek and Emlen, 1999; Hunt and Simmons,2000). <strong>The</strong> horns are used <strong>in</strong> <strong>contest</strong> <strong>competition</strong> with other <strong>male</strong>s and large horn size is associated with <strong>in</strong>creasedreproductive success. Hornless <strong>male</strong>s seek mat<strong>in</strong>g with fe<strong>male</strong>s <strong>in</strong>side tunnels while avoid<strong>in</strong>g aggressive <strong>in</strong>teractionswith guard<strong>in</strong>g <strong>male</strong>s (Mozek and Emlen, 2000). <strong>The</strong> two reproductive tactics used by <strong>male</strong> dung beetles reflectphenotypic plasticity but may represent the first step towards genetically determ<strong>in</strong>ed polymorphism. <strong>The</strong>re are severalreported cases <strong>of</strong> genetically determ<strong>in</strong>ed polymorphism (Zimmerer and Kallman, 1989; Shuster and Wade, 1991; Lank5
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>et al., 1995; S<strong>in</strong>ervo and Lively, 1996; Tuttle, 2003; Hurtado-Gonzales and Uy, 2009), which may <strong>in</strong> turn represent afirst step towards <strong>speciation</strong>.In conclusion, the <strong>role</strong> <strong>of</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>in</strong> promot<strong>in</strong>g and ma<strong>in</strong>ta<strong>in</strong><strong>in</strong>g variation among and betweenpopulations rema<strong>in</strong>s a modestly explored scientific area. S<strong>in</strong>ce aggression from competitors is considered to be onemajor cost associated with hav<strong>in</strong>g many competitors with similar fight<strong>in</strong>g ability, <strong>male</strong> <strong>contest</strong> <strong>competition</strong> may result<strong>in</strong> disruptive selection favor<strong>in</strong>g <strong>male</strong>s with extreme strategies (i.e. <strong>of</strong>ten a dom<strong>in</strong>ant versus a sneak<strong>in</strong>g strategy) andnegative frequency-dependent selection on discrete morphs. <strong>The</strong>re are many possible directions for future research butwe would like to p<strong>in</strong>po<strong>in</strong>t three questions that we f<strong>in</strong>d particularly <strong>in</strong>terest<strong>in</strong>g. While it is <strong>in</strong>tuitive that sexually selectedtraits directly used <strong>in</strong> combat (e.g. horns) are associated with dom<strong>in</strong>ant strategies, a rema<strong>in</strong><strong>in</strong>g question is to what extentthere are pre-exist<strong>in</strong>g physiological biases mak<strong>in</strong>g certa<strong>in</strong> colors more likely to become associated with dom<strong>in</strong>antstrategies (e.g. red color may be perceived as more threaten<strong>in</strong>g than blue color). Another unexplored area <strong>of</strong> research isto <strong>in</strong>vestigate the possible <strong>role</strong> <strong>of</strong> learn<strong>in</strong>g from previous <strong>in</strong>teractions with competitors <strong>in</strong> re<strong>in</strong>forc<strong>in</strong>g asymmetries <strong>in</strong>aggression and/or <strong>in</strong> the outcome <strong>of</strong> <strong>contest</strong>s. F<strong>in</strong>ally, more research aimed at reveal<strong>in</strong>g the costs and benefits associatedwith different strategies (i.e. <strong>in</strong>clud<strong>in</strong>g differences <strong>in</strong> whole suites <strong>of</strong> <strong>male</strong> traits that <strong>in</strong>fluence fight<strong>in</strong>g ability orsneak<strong>in</strong>g ability) and how these depend on the social and ecological environment is needed.3 <strong>The</strong> Ecological ContextStable co-existence between species is facilitated by differences <strong>in</strong> niche use. A relevant question then becomes howmay divergence <strong>in</strong> sexually selected traits used <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>in</strong>fluence niche use and vice versa? When<strong>male</strong> <strong>contest</strong> <strong>competition</strong> causes negative frequency-dependent selection, it will facilitate the ma<strong>in</strong>tenance <strong>of</strong>polymorphism with little divergence <strong>in</strong> niche use. Males with more different sexually selected traits will be able to usemore similar niches under sympatric conditions (see example <strong>of</strong> cichlids above).When divergence <strong>in</strong> traits used <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> is associated with differences <strong>in</strong> aggression and/orfight<strong>in</strong>g ability, this will also <strong>in</strong>fluence access to other resources. Aggressive <strong>in</strong>terspecific <strong>in</strong>terference has for examplebeen shown to drive divergence <strong>in</strong> habitat use by two species <strong>of</strong> land snails: Euhadra quaesita and E. peliomphala.Members <strong>of</strong> the less aggressive species, E. quaesita, are more terrestrial when they live <strong>in</strong> allopatry as compared towhen they live <strong>in</strong> sympatry with E. peliomphala (Kimura and Chiba, 2010). In general, when <strong>male</strong>s fight <strong>over</strong> resourcesnecessary to attract fe<strong>male</strong>s, the outcome <strong>of</strong> such <strong>contest</strong>s <strong>of</strong>ten determ<strong>in</strong>es <strong>in</strong> what type <strong>of</strong> environment their <strong>of</strong>fspr<strong>in</strong>gare be<strong>in</strong>g raised <strong>in</strong>. Habitat impr<strong>in</strong>t<strong>in</strong>g can then promote the association between competitive strategy and habitat useacross generations (see e.g. Vall<strong>in</strong> and Qvarnström 2011).An association between divergence <strong>in</strong> sexually selected traits used <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> and divergence <strong>in</strong>niche use is also expected because the costs and benefits <strong>of</strong> us<strong>in</strong>g a dom<strong>in</strong>ant versus less competitive strategy shouldvary across ecological contexts. To understand population divergence <strong>in</strong> <strong>male</strong> traits used <strong>in</strong> <strong>contest</strong> <strong>competition</strong> <strong>over</strong>access to <strong>mates</strong> a wide range <strong>of</strong> costs associated with possess<strong>in</strong>g such traits needs to be taken <strong>in</strong>to account. Individuals<strong>in</strong> allopatric populations most likely experience different selective regimes due to differences <strong>in</strong> the ecological andsocial environment, which <strong>in</strong> turn result <strong>in</strong> differences <strong>in</strong> the adaptive optima <strong>of</strong> traits used <strong>in</strong> <strong>contest</strong> <strong>competition</strong> (Fig.3). It has been suggested that the level <strong>of</strong> sexual <strong>competition</strong> should be higher <strong>in</strong> stable and resource rich environmentswhile harsh environments limit sexual selection, which seems <strong>in</strong>tuitive s<strong>in</strong>ce there is a high cost <strong>of</strong> <strong>contest</strong> behavior(Briffa and Sneddon, 2007; Chellappa and Hunt<strong>in</strong>gford, 1989). However, under stressful conditions, where the lifeexpectancy is lower, an evolutionary switch to higher allocation <strong>of</strong> resources to <strong>in</strong>crease reproductive success is alsoknown to occur (Birkhead et al., 1999; Polak and Starmer, 1998). Thus, it may be hard to predict the direction <strong>of</strong> theevolutionary response <strong>of</strong> traits <strong>in</strong>volved <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>in</strong> stressful environments.6
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>One important stress factor that <strong>of</strong>ten varies considerably across both time and space is parasites. Aggressivebehavior is known to be associated with high testosterone levels (e.g. Mart<strong>in</strong>ez-Sanchis et al., 2003) and a majorphysiological cost <strong>of</strong> high testosterone levels is impaired immune function. Several elegant studies have been performed<strong>in</strong> specific populations. Testosterone implants have been shown to <strong>in</strong>crease parasite <strong>in</strong>fection <strong>in</strong> red grouse, for example(Mougeot et al., 2004, 2005). Elevated levels <strong>of</strong> testosterone may also result <strong>in</strong> higher exposure and transmission <strong>of</strong>parasites through changes <strong>in</strong> behavior (Grear et al., 2009). Thus, aggressive behavior <strong>in</strong> itself may lead to higherexposure to parasites. In any case, the ma<strong>in</strong> prediction would be that high levels <strong>of</strong> parasites <strong>in</strong> the environment shouldselect for reduced levels <strong>of</strong> aggression.Another commonly discussed negative effect <strong>of</strong> secondary sexual traits, which is also likely to differ betweenenvironments and populations, is the <strong>in</strong>creased risk <strong>of</strong> predation. Engagement <strong>in</strong> <strong>in</strong>trasexual <strong>contest</strong>s has been shown to<strong>in</strong>crease predation risk (e.g. Jakobsson, 1995; Dunn, 2004), and thus, predation is assumed to limit both the degree <strong>of</strong><strong>in</strong>trasexual <strong>contest</strong> <strong>competition</strong> and the degree <strong>of</strong> <strong>in</strong>tersexual signal<strong>in</strong>g. <strong>The</strong> relationship between predation pressure andaggressive behaviour has recently been studied <strong>in</strong> Strawberry poison frogs Dendrobates pumilio. Populations <strong>of</strong> thisspecies <strong>in</strong> the Archipelago <strong>of</strong> Bocas del Toro <strong>in</strong> northwestern Panama have ga<strong>in</strong>ed special attention because <strong>of</strong> theirgreat divergence <strong>in</strong> body color (Summers et al., 2003; Siddiqi et al., 2004) and because the color present <strong>in</strong> this speciesis shown to be <strong>in</strong>volved <strong>in</strong> predator <strong>in</strong>teractions (Summers and Clough, 2001; Saporito et al., 2007). Populations rangefrom cryptic to highly conspicuous (Wang, 2008, Fig. 2). Strawberry poison frogs show strong territoriality with highlevels <strong>of</strong> aggression (Pröhl, 2005) but the ma<strong>in</strong> focus <strong>in</strong> studies on <strong>in</strong>teractions between natural and sexual selection <strong>in</strong>this species has been directed to fe<strong>male</strong> mate choice (Summers et al. 1999; Reynolds and Fitzpatrick, 2007; Maan andCumm<strong>in</strong>gs, 2008). Based on the hypothesis that coloration differences <strong>in</strong> D.pumilio reflect predation avoidancestrategies, Rudh et al (2011) argued that aposematic <strong>in</strong>dividuals should be relieved from the behavioral constra<strong>in</strong>ts thatcryptic <strong>in</strong>dividuals suffer and that the evolution <strong>of</strong> aposematic coloration could therefore be facilitated through the jo<strong>in</strong>taction <strong>of</strong> natural and sexual selection. <strong>The</strong> authors suggest that due to the loss <strong>of</strong> aposematic coloration <strong>in</strong> thecryptically colored populations, the predation constra<strong>in</strong>t on aggressive behavior is restored, which causes lower levels<strong>of</strong> behaviors that affect exposure to predators. This idea was supported by the f<strong>in</strong>d<strong>in</strong>g that <strong>male</strong>s <strong>in</strong> more conspicuouspopulations <strong>of</strong> D.pumilio show higher levels <strong>of</strong> exposure while call<strong>in</strong>g (Rudh et al., 2011). Male call<strong>in</strong>g <strong>in</strong> this speciesserves both as an <strong>in</strong>tersexual signal to attract fe<strong>male</strong>s and as an <strong>in</strong>trasexual aggressive signal to defend territories. In astudy by Crothers et al (2011), <strong>male</strong>s from an island population with aposematic red <strong>in</strong>dividuals were tested for<strong>in</strong>traspecific aggression. Us<strong>in</strong>g two-way experimental setups, the authors found that not only did more brightly colored<strong>male</strong>s act more aggressively <strong>in</strong> general, but there was also a higher level <strong>of</strong> aggression directed towards <strong>male</strong>s withartificially <strong>in</strong>creased brightness. <strong>The</strong>se f<strong>in</strong>d<strong>in</strong>gs suggest that even small differences <strong>in</strong> color stimuli may have significanteffects on <strong>male</strong> aggression levels. Rudh et al (manuscript) compared <strong>male</strong>-<strong>male</strong> aggression <strong>in</strong> several populations witheither cryptic or conspicuous <strong>in</strong>dividuals. <strong>The</strong>ir results showed a higher level <strong>of</strong> aggression <strong>of</strong> <strong>in</strong>dividuals belong<strong>in</strong>g toconspicuous populations than <strong>in</strong> <strong>in</strong>dividuals from cryptic populations.To conclude, there are many reasons why adaptation to different environments should commonly co<strong>in</strong>cide withsome divergence <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> ability and vice versa. Males us<strong>in</strong>g dom<strong>in</strong>ant strategies <strong>of</strong>ten behave moreaggressively, have higher testosterone levels and are <strong>in</strong> many cases larger than <strong>male</strong>s us<strong>in</strong>g a sneaker strategy (Gross,1996). <strong>The</strong> costs and benefits associated with these various traits are known to vary across ecological contexts.Although costs associated with <strong>male</strong> <strong>contest</strong> <strong>competition</strong> are documented <strong>in</strong> the literature, surpris<strong>in</strong>gly little researchhas been performed on how such costs <strong>in</strong>fluence population divergences and the build up <strong>of</strong> reproductive isolation.Exam<strong>in</strong><strong>in</strong>g how divergence <strong>in</strong> traits used <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>in</strong>fluence the build up <strong>of</strong> reproductive isolationtherefore provides a major challenge for future studies.7
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>4 Male Contest Competition and Character Displacement<strong>The</strong> importance <strong>of</strong> disruptive selection caused by <strong>competition</strong> <strong>over</strong> limit<strong>in</strong>g resources (Dieckmann and Doebeli,1999) is not limited to the situation <strong>of</strong> <strong>speciation</strong> under constant sympatric conditions. S<strong>in</strong>ce the development <strong>of</strong>complete reproductive isolation is <strong>of</strong>ten a slow evolutionary process, many populations that start diverg<strong>in</strong>g <strong>in</strong> allopatryspend at least some time at secondary contact zones before they have developed complete reproductive isolation.Harmful <strong>in</strong>terspecific <strong>in</strong>teractions such as <strong>competition</strong> <strong>over</strong> <strong>mates</strong> and/or resources can be reduced by divergence <strong>in</strong>resource-use or <strong>in</strong> reproductive phenotypes, i.e. character displacement (Brown and Wilson, 1956). Characterdisplacement is <strong>of</strong>ten further divided <strong>in</strong>to reproductive character displacement (character displacement <strong>in</strong> traitsassociated with reproduction) and ecological character displacement (character displacement <strong>in</strong> traits associated withresource use). As character displacement favors the evolution <strong>of</strong> novel resource use or reproductive traits, it can both<strong>in</strong>itiate and f<strong>in</strong>alize the <strong>speciation</strong> process and is potentially a lead<strong>in</strong>g cause <strong>of</strong> adaptive diversification (Pfennig andPfennig, 2009; Schluter, 2000). Displacement <strong>of</strong> reproductive characters through heterospecific <strong>competition</strong> between<strong>male</strong>s is likely when; a) the level <strong>of</strong> pre-mat<strong>in</strong>g isolation is low and <strong>male</strong>s belong<strong>in</strong>g to the two different populationscompete <strong>over</strong> fe<strong>male</strong>s, b) fe<strong>male</strong>s <strong>of</strong> both populations use the same types <strong>of</strong> resources for their reproduction, which<strong>male</strong>s <strong>in</strong> turn compete <strong>over</strong>, or c) when there is misplaced aggression towards heterospecific <strong>male</strong>s that resembleconspecific <strong>male</strong>s and therefore are perceived as competitors (i.e. reproductive <strong>in</strong>terference, Grön<strong>in</strong>g and Hochkirch,2008). Character displacement at secondary contact is facilitated if the characters <strong>in</strong>volved have already to some extentdiverged between the two populations <strong>in</strong> allopatry (Milligan, 1985; Schluter, 2000). Abundant stand<strong>in</strong>g variation alsopromotes character displacement (Pfennig and Pfennig, 2009), and further divergence <strong>in</strong> sympatry can then proceedrapidly as selection acts on phenotypes already present. Another general pattern is that the strength <strong>of</strong> selection to avoidreproductive <strong>in</strong>teractions at secondary contact <strong>of</strong>ten differs between the two populations, result<strong>in</strong>g <strong>in</strong> asymmetriccharacter displacement. Such asymmetries can be <strong>in</strong>terpreted as a trade-<strong>of</strong>f between the benefits <strong>of</strong> avoid<strong>in</strong>g<strong>competition</strong> with heterospecifics and the costs <strong>of</strong> hav<strong>in</strong>g a displaced character (Cooley, 2007; Pfennig and Pfennig,2009).Several elegant studies have exam<strong>in</strong>ed the <strong>role</strong> <strong>of</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>in</strong> driv<strong>in</strong>g reproductive characterdisplacement <strong>in</strong> damselflies. Male damselflies defend small mat<strong>in</strong>g territories close to open water (Alcock, 1987). Acompell<strong>in</strong>g example <strong>of</strong> asymmetric reproductive character displacement driven by heterospecific <strong>male</strong> <strong>competition</strong> isthe w<strong>in</strong>g spot <strong>of</strong> the damselfly Calopteyrx splendens <strong>The</strong> size <strong>of</strong> the w<strong>in</strong>g spots are smaller <strong>in</strong> <strong>male</strong>s belong<strong>in</strong>g to C.splendens populations that are liv<strong>in</strong>g <strong>in</strong> sympatry with the dom<strong>in</strong>ant C. virgo, which display higher levels <strong>of</strong> aggressiontowards C. splendens with large black spots (Tynkkynen et al., 2004; 2005; 2006; Fig. 2). Selection aga<strong>in</strong>sthybridization provides a non-mutually exclusive additional explanation for the observed pattern s<strong>in</strong>ce w<strong>in</strong>g colorationalso functions as a cue <strong>in</strong> mate choice (Svensson et al., 2007). Anderson and Grether (2010) focused on the effect <strong>of</strong>competitor recognition failure <strong>in</strong> Hetaer<strong>in</strong>a damselflies by manipulat<strong>in</strong>g conspecific <strong>in</strong>truders to resembleheterospecific <strong>male</strong>s. This treatment lowered the aggressive response <strong>of</strong> territorial <strong>male</strong>s <strong>in</strong> areas where suchheterospecific <strong>male</strong>s naturally occurred, while no reduction <strong>in</strong> aggression was observed <strong>in</strong> allopatric sites. <strong>The</strong>seexperiments imply an <strong>in</strong>creased ability to avoid heterospecific aggression <strong>in</strong> sympatric populations. An excit<strong>in</strong>gquestion is to what extent variation <strong>in</strong> competitor recognition (i.e. conspecific <strong>male</strong>s) is a genetically determ<strong>in</strong>ed trait orreflects learn<strong>in</strong>g. S<strong>in</strong>ce the relative species abundance changes with<strong>in</strong> sites (Anderson and Grether 2010), learn<strong>in</strong>g fromprevious experiences is likely to be important.<strong>The</strong> relationship between reproductive character displacement and ecological divergence has recently beenexam<strong>in</strong>ed <strong>in</strong> a young Ficedula flycatcher hybrid zone. Collared flycatchers colonized the Swedish island <strong>of</strong> Öland <strong>in</strong> theBaltic Sea about 50 years ago and are now rapidly exclud<strong>in</strong>g their pied flycatcher predecessors from the preferred8
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>breed<strong>in</strong>g sites (Qvarnström et al., 2009). <strong>The</strong> ma<strong>in</strong> mechanism driv<strong>in</strong>g this exclusion is that young <strong>male</strong> pied flycatchersfail to establish breed<strong>in</strong>g territories as the density <strong>of</strong> <strong>male</strong> collared flycatchers <strong>in</strong>creases (Vall<strong>in</strong> et al., 2011a), result<strong>in</strong>g<strong>in</strong> a shift <strong>in</strong> the breed<strong>in</strong>g habitats <strong>of</strong> pied flycatchers from deciduous forest towards mixed or coniferous forest (Vall<strong>in</strong> etal., 2011b). Because the peak <strong>in</strong> food abundance differs between habitats, the spatial segregation is paralleled by an<strong>in</strong>creased divergence <strong>in</strong> tim<strong>in</strong>g <strong>of</strong> breed<strong>in</strong>g between the two species (Vall<strong>in</strong> et al., 2011b). Male pied flycatchers varyfrom black and white to brown and white throughout their range (Fig. 2), but the brown coloration is more common <strong>in</strong>areas where they co-occur with the black and white collared flycatchers (Drost, 1936; Roskaft and Järvi, 1992; Sætre etal., 1993; Alatalo et al., 1994; Huhta et al., 1997). Previous experiments have shown that <strong>in</strong>terspecific aggression isrelaxed for browner <strong>male</strong> pied flycatchers (Král et al., 1988; Gustafsson and Pärt, 1991; Sætre et al., 1993), and brown<strong>male</strong> pied flycatchers experienced higher relative fitness than black <strong>male</strong>s when faced with heterospecific <strong>competition</strong>(Vall<strong>in</strong> et al., 2011b). More<strong>over</strong>, relatively brown <strong>male</strong> pied flycatchers were found to settle <strong>in</strong> habitats most differentfrom the ones defended by collared flycatchers and to breed relatively late (Vall<strong>in</strong> et al., 2011b). Thus, pied flycatcherswith the most divergent breed<strong>in</strong>g strategy compared to collared flycatchers appear to be favored by selection <strong>in</strong> thehybrid zone. However, because divergent plumage characters promote local co-existence between the two flycatcherspecies, it <strong>in</strong>creases the risk <strong>of</strong> hybridization <strong>in</strong> this young hybrid zone (Vall<strong>in</strong> et al., 2011b). <strong>The</strong> f<strong>in</strong>d<strong>in</strong>gs from thestudies on the Ficedula flycatchers lead to the idea that <strong>in</strong>terspecific <strong>competition</strong> among <strong>male</strong>s <strong>over</strong> resources necessaryto attract fe<strong>male</strong>s (i.e. suitable nest sites <strong>in</strong> high quality habitat) may <strong>in</strong> general give rise to a feedback loop betweenecological divergence and reproductive character displacement (Fig. 4; Vall<strong>in</strong> et al., 2011b). This is because ecologicaldivergence <strong>in</strong> allopatry is likely to <strong>in</strong>fluence aggressive competitive abilities. At secondary contact, the subdom<strong>in</strong>antspecies is displaced <strong>in</strong>to a poorer habitat, which <strong>in</strong> turn leads to further divergence <strong>in</strong> competitive ability.In conclusion, although most studies <strong>of</strong> reproductive character displacement have focused on the <strong>role</strong> <strong>of</strong>re<strong>in</strong>forcement, i.e. selection aga<strong>in</strong>st hybridization driv<strong>in</strong>g divergence (Noor, 1995; Sætre et al., 1997; Nosil et al., 2003;Hosk<strong>in</strong> et al., 2005) there is an <strong>in</strong>creas<strong>in</strong>g awareness that other species <strong>in</strong>teractions may also cause this pattern (Hosk<strong>in</strong>and Higgie, 2010). In particular, the importance <strong>of</strong> <strong>in</strong>terspecific aggression is ga<strong>in</strong><strong>in</strong>g <strong>in</strong>creas<strong>in</strong>g support (Seehausenand Schluter, 2004; Tynkkynen et al., 2004; 2005; Grether et al., 2009; Anderson and Grether, 2010, Vall<strong>in</strong> et al.,2011b). However, these two selective processes are not mutually exclusive and we would like to encourage futureresearch on their possible <strong>in</strong>teraction.5 How Can Male Contest Competition <strong>over</strong> Mates Influence the Evolution <strong>of</strong>Reproductive Isolation?Speciation is def<strong>in</strong>ed as a split <strong>of</strong> one species <strong>in</strong>to two or more new species that follow <strong>in</strong>dependent evolutionarypathways (Dobzhansky, 1935). In sexually reproduc<strong>in</strong>g organisms, the development <strong>of</strong> reproductive isolation is aprerequisite for the occurrence <strong>of</strong> this process. <strong>The</strong>refore, f<strong>in</strong>d<strong>in</strong>g out how population divergence through naturalselection, sexual selection and/or genetic drift leads to the build up <strong>of</strong> reproductive isolation lies at the heart <strong>of</strong> researchon the mechanisms <strong>of</strong> <strong>speciation</strong>. <strong>The</strong> two most straightforward pathways to population divergence driven by <strong>male</strong><strong>contest</strong> <strong>competition</strong> is that there is disruptive selection <strong>in</strong> sympatry i.e. hybrids receive a disproportionate amount <strong>of</strong>attack from both species (Fig. 1), or divergence <strong>in</strong> allopatry caused by differences <strong>in</strong> the balance between the costs andbenefits associated with different <strong>male</strong> competitive tactics across different ecological conditions (Fig. 3). S<strong>in</strong>ce <strong>male</strong><strong>contest</strong> abilities <strong>of</strong>ten <strong>in</strong>clude a number <strong>of</strong> behavioral, morphological and physiological traits, there may be strongselection aga<strong>in</strong>st hybrids and backcrossed <strong>in</strong>dividuals with mismatched trait comb<strong>in</strong>ations.9
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong><strong>The</strong>re are several potential evolutionary pathways to reproductive isolation when there is population divergence <strong>in</strong><strong>male</strong> traits associated with fight<strong>in</strong>g abilities <strong>in</strong> resource based breed<strong>in</strong>g systems. Apart from caus<strong>in</strong>g post-zygoticreproductive isolation through selection aga<strong>in</strong>st hybrids, divergence <strong>in</strong> ability to compete <strong>over</strong> resources used to attractfe<strong>male</strong>s can also lead to pre-zygotic reproductive isolation as a by-product, e.g. through displacement <strong>in</strong> habitat choice(e.g. Feder et al., 1994; Schluter, 2000) or breed<strong>in</strong>g time (Théron and Combes,1995; Hendry and Day, 2005). Anexample <strong>of</strong> habitat segregation driven by <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>over</strong> nest sites is provided by the flycatcher examplementioned above. Habitat segregation has <strong>in</strong>creased reproductive isolation between the two flycatcher species (Vall<strong>in</strong>and Qvarnström, 2011); <strong>male</strong> pied flycatchers that are displaced <strong>in</strong>to the less preferred coniferous habitat are also lesslikely to hybridize. A cross-foster<strong>in</strong>g experiment <strong>of</strong> nestl<strong>in</strong>gs and eggs between the two flycatcher species suggestedthat learned habitat choice further strengthens the barrier between the two species (Vall<strong>in</strong> and Qvarnström, 2011). Al<strong>in</strong>k between divergence <strong>in</strong> coloration, level <strong>of</strong> aggression and segregation between call<strong>in</strong>g and/or breed<strong>in</strong>g sites hasbeen suggested to <strong>in</strong>crease the likelihood <strong>of</strong> assortative mat<strong>in</strong>g <strong>in</strong> D. pumilio (Rudh et al., 2011). Thus, divergence <strong>in</strong>traits used <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> can lead to both pre- and post- zygotic reproductive isolation and these traitsmay hence sometimes function as “magic traits” (Gavrilets, 2004; Servedio et al., 2011).Another pathway to reproductive isolation is if divergence <strong>in</strong> traits used <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> is l<strong>in</strong>ked withdivergence <strong>in</strong> fe<strong>male</strong> choice based on these traits. <strong>The</strong>re are several comprehensive review articles on the <strong>in</strong>teractionbetween <strong>male</strong> <strong>contest</strong> <strong>competition</strong> and fe<strong>male</strong> choice <strong>in</strong> caus<strong>in</strong>g evolution through sexual selection (e.g. Berglund et al.,1996; Qvarnström and Forsgren, 1998; Wong and Candol<strong>in</strong>, 2005; Hunt et al., 2009). However, there has been a lack <strong>of</strong>attempts to <strong>in</strong>vestigate how the <strong>in</strong>teraction between these two mechanisms <strong>of</strong> sexual selection may <strong>in</strong>fluence the<strong>speciation</strong> process. Below we outl<strong>in</strong>e a few possible scenarios, but we would like to emphasize that there is too littleresearch done <strong>in</strong> this field to draw major conclusions.In some species, e.g. with large sexual size dimorphism, <strong>male</strong> <strong>contest</strong> <strong>competition</strong> has the largest impact on pair<strong>in</strong>gpatterns, leav<strong>in</strong>g little or no scope for fe<strong>male</strong>s to freely exert mate choice. S<strong>in</strong>ce the costs <strong>of</strong> heterospecific mat<strong>in</strong>g <strong>of</strong>tenare expected to be higher for fe<strong>male</strong>s than for <strong>male</strong>s, fe<strong>male</strong>s should develop species recognition abilities faster than<strong>male</strong>s (Wirtz, 1999). As a consequence, heterospecific pair<strong>in</strong>gs may be more common between young species whenfe<strong>male</strong>s are unable to exert free choice. In resource based breed<strong>in</strong>g systems, <strong>male</strong> <strong>competition</strong> <strong>of</strong>ten determ<strong>in</strong>es amongwhich <strong>male</strong>s fe<strong>male</strong>s can choose, e.g. mak<strong>in</strong>g the two mechanisms <strong>of</strong> sexual selection work <strong>in</strong> a sequential manner. If<strong>male</strong> <strong>competition</strong> results <strong>in</strong> a sort<strong>in</strong>g <strong>of</strong> <strong>male</strong>s <strong>of</strong> different species <strong>in</strong>to different habitat types, it may facilitate fe<strong>male</strong>choice <strong>of</strong> conspecific <strong>male</strong>s (e.g. Vall<strong>in</strong> and Qvanström, 2011).When both <strong>male</strong> <strong>competition</strong> and fe<strong>male</strong> choice occur (<strong>in</strong> sequence or <strong>in</strong> parallel), a relevant question becomeswhether the two mechanisms <strong>of</strong> sexual selection will work <strong>in</strong> the same direction on the <strong>male</strong> traits. As mentioned <strong>in</strong> the<strong>in</strong>troduction, fe<strong>male</strong> choice is generally assumed to operate <strong>in</strong> a unidirectional manner. Does this mean that fe<strong>male</strong>choice should act aga<strong>in</strong>st population divergence <strong>in</strong> sexually selected traits? Divergence <strong>in</strong> traits used <strong>in</strong> <strong>contest</strong><strong>competition</strong> could quickly lead to population assortative mat<strong>in</strong>g if sexual impr<strong>in</strong>t<strong>in</strong>g has a large impact on pair<strong>in</strong>gpatterns. Sexual impr<strong>in</strong>t<strong>in</strong>g means that <strong>in</strong>dividuals learn characteristics <strong>of</strong> their parents that enable them to f<strong>in</strong>d asuitable mate dur<strong>in</strong>g adulthood (e.g. Bateson, 1966; Clayton, 1993). One reason why sexual impr<strong>in</strong>t<strong>in</strong>g may play animportant <strong>role</strong> <strong>in</strong> <strong>speciation</strong> is that it works as a one-allele mechanism. <strong>The</strong> same allele <strong>in</strong> each <strong>of</strong> the divergedpopulations promotes assortative mat<strong>in</strong>g based on the traits that separate the two populations. This means thatdevelopment <strong>of</strong> assortative mat<strong>in</strong>g is not sensitive to gene flow (Irw<strong>in</strong> and Price, 1999; Servedio, 2009). However, thistype <strong>of</strong> one-allele mechanism cannot expla<strong>in</strong> population divergence <strong>in</strong> the first place (Servedio, 2009), mak<strong>in</strong>g thecomb<strong>in</strong>ation <strong>of</strong> this mechanism and population divergence through <strong>male</strong> <strong>contest</strong> <strong>competition</strong> a particularly goodcandidate as an eng<strong>in</strong>e <strong>of</strong> <strong>speciation</strong>. Cross-foster<strong>in</strong>g experiments us<strong>in</strong>g birds liv<strong>in</strong>g <strong>in</strong> natural populations have shown10
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>that exposure to heterospecific parents dur<strong>in</strong>g the nestl<strong>in</strong>g stage may result <strong>in</strong> a sexual preference for heterospecific<strong>in</strong>dividuals when adult (Harris, 1970; Fabricius, 1991), imply<strong>in</strong>g high importance for sexual impr<strong>in</strong>t<strong>in</strong>g <strong>in</strong> assortativemat<strong>in</strong>g under normal conditions. Recently, sexual impr<strong>in</strong>t<strong>in</strong>g was also found to promote assortative mat<strong>in</strong>g betweenbenthic and limnetic sticklebacks (Kozak et al., 2011).Intensive <strong>male</strong> <strong>contest</strong> <strong>competition</strong> may lead to the evolution <strong>of</strong> traits that <strong>in</strong>crease <strong>male</strong> mat<strong>in</strong>g success but reducefe<strong>male</strong> fitness, which <strong>in</strong> turn may trigger an evolutionary arms race between the two sexes whereby fe<strong>male</strong>s evolvecounter adaptations (Arnqvist and Rowe, 2002). Population divergence driven by such arms races may lead toreproductive isolation as a side effect. Sexual conflict is def<strong>in</strong>ed as a conflict between the evolutionary <strong>in</strong>terests <strong>of</strong> thetwo sexes (Parker, 1979) and can be further classified <strong>in</strong>to <strong>in</strong>tra- or <strong>in</strong>terlocus sexual conflicts. An <strong>in</strong>tralocus sexualconflict arises when a particular allele <strong>in</strong>creases fitness when expressed <strong>in</strong> <strong>in</strong>dividuals belong<strong>in</strong>g to one <strong>of</strong> the two sexesbut decreases fitness when expressed <strong>in</strong> <strong>in</strong>dividuals belong<strong>in</strong>g to the other sex (Rice and Chippendale, 2001). An<strong>in</strong>terlocus sexual conflict arises when an adaptation <strong>in</strong> <strong>male</strong>s causes reduced fitness <strong>in</strong> fe<strong>male</strong>s and fe<strong>male</strong>s respond byevolv<strong>in</strong>g counter-adaptations. This latter situation may result <strong>in</strong> co-evolutionary arms-races between the two sexes<strong>in</strong>volv<strong>in</strong>g genes at different loci <strong>in</strong> each sex (Arnquist and Rowe, 2005). Genes underly<strong>in</strong>g <strong>male</strong> traits used <strong>in</strong> <strong>contest</strong><strong>competition</strong> <strong>over</strong> fe<strong>male</strong>s are likely to reduce fitness when expressed <strong>in</strong> fe<strong>male</strong>s. <strong>The</strong> <strong>in</strong>ter-sexual genetic correlation forfitness may therefore be expected to be lower <strong>in</strong> species with more extreme sex-<strong>role</strong>s (Qvarnström et al., 2006). Anevolutionary solution to this situation is that genes underly<strong>in</strong>g traits used <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong>, such as largebody size and horns, are only expressed <strong>in</strong> <strong>male</strong>s result<strong>in</strong>g <strong>in</strong> pronounced phenotypic sexual dimorphisms. Evolutiondriven by <strong>male</strong> <strong>contest</strong> <strong>competition</strong> may also cause a number <strong>of</strong> <strong>in</strong>ter-locus sexual conflicts. Apart from trigger<strong>in</strong>g tight<strong>in</strong>tersexual co-evolutionary arms-races, <strong>in</strong>creased adaptation to <strong>contest</strong> <strong>competition</strong> may occur at the expense <strong>of</strong> othertraits <strong>in</strong> <strong>male</strong>s, such as reduced paternal care. Fe<strong>male</strong>s may, <strong>in</strong> turn, be selected to reduce their litter/clutch size. Sexualconflict aris<strong>in</strong>g through adaptations related to <strong>male</strong> <strong>contest</strong> <strong>competition</strong> should be expected to cause fast evolutionarychanges <strong>in</strong> the genetic regulation <strong>of</strong> traits (reflected by sexual size dimorphisms) and <strong>in</strong> whole suites <strong>of</strong> traits <strong>in</strong> bothsexes. <strong>The</strong> likelihood <strong>of</strong> evolution <strong>of</strong> reproductive isolation between populations experienc<strong>in</strong>g different levels <strong>of</strong> <strong>male</strong><strong>contest</strong> <strong>competition</strong> should therefore be relatively high. <strong>The</strong>oretical model<strong>in</strong>g has shown that sexual conflict canpromote <strong>speciation</strong> by lead<strong>in</strong>g to fast changes <strong>in</strong> traits <strong>in</strong>volved <strong>in</strong> reproduction such as egg-sperm prote<strong>in</strong>s (e.g.Gavrilets, 2000), but there should be many more unexplored pathways to reproductive isolation driven by sexualconflict.<strong>The</strong> <strong>role</strong> <strong>of</strong> <strong>in</strong>tr<strong>in</strong>sic genetic <strong>in</strong>compatibilities <strong>in</strong> the process <strong>of</strong> <strong>speciation</strong> is debated. This is because the evolution<strong>of</strong> hybrid sterility is a slow process compared to the rate <strong>of</strong> <strong>speciation</strong> <strong>in</strong> several taxa (Price and Bouvier, 2002;Mendelson, 2003). Can <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>in</strong>fluence the speed by which genetic <strong>in</strong>compatibilities accumulate?Intr<strong>in</strong>sic sources <strong>of</strong> hybrid dysfunction are assumed to arise through epistatic <strong>in</strong>teractions between genes from differentgenomes (Dobzhansky, 1940; Muller, 1940). A general pattern is that there is a greater fitness reduction <strong>in</strong> hybrids <strong>of</strong>the heterogametic sex, i.e. Haldane’s rule (Haldane, 1922). <strong>The</strong> faster <strong>male</strong> hypothesis builds on the idea that Haldane’srule is a consequence <strong>of</strong> <strong>in</strong>tensive sexual selection on <strong>male</strong>s lead<strong>in</strong>g to a faster <strong>in</strong>compatibility <strong>of</strong> <strong>male</strong> limited traits(Wu and Davies, 1993; Wu et al., 1996). <strong>The</strong>se <strong>male</strong> traits are primary sexual traits and not secondary sexual traits suchas horns and spears that have a direct function <strong>in</strong> <strong>contest</strong> <strong>competition</strong>. However, <strong>in</strong>vestigat<strong>in</strong>g the <strong>role</strong> <strong>of</strong> <strong>male</strong> <strong>contest</strong><strong>competition</strong> <strong>in</strong> driv<strong>in</strong>g divergence <strong>in</strong> <strong>male</strong> competitive tactics (which <strong>of</strong>ten also <strong>in</strong>clude differences <strong>in</strong> primary sexualtraits such as sperm characteristics) would be highly relevant. Recent developments <strong>in</strong> genomics and high-throughputtechnologies are open<strong>in</strong>g novel possibilities to reveal l<strong>in</strong>ks between divergence between natural populations and thebuild up <strong>of</strong> genetic <strong>in</strong>compatibilities (Rice et al., 2011).11
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>In conclusion, there are many potential ways <strong>in</strong> which <strong>male</strong> <strong>contest</strong> <strong>competition</strong> can <strong>in</strong>fluence populationdivergence and the build up <strong>of</strong> reproductive isolation. For example, <strong>in</strong> resource based breed<strong>in</strong>g systems, pre-zygoticisolation through habitat segregation may arise as a side effect <strong>of</strong> biases <strong>in</strong> <strong>male</strong> competitive abilities betweenpopulations. <strong>The</strong>re are however also possible ways by which <strong>male</strong> <strong>contest</strong> <strong>competition</strong> can prevent pre-zygotic isolation.Male <strong>contest</strong> <strong>competition</strong> can sometimes limit fe<strong>male</strong> ability to choose conspecific <strong>mates</strong>. How the two ma<strong>in</strong>mechanisms <strong>of</strong> sexual selection (i.e. <strong>male</strong> <strong>contest</strong> <strong>competition</strong> and fe<strong>male</strong> choice) <strong>in</strong>teract dur<strong>in</strong>g the <strong>speciation</strong> process<strong>in</strong> general rema<strong>in</strong>s a largely unexplored research area. We would also like to encourage future research on the <strong>role</strong> <strong>of</strong><strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>in</strong> the evolution <strong>of</strong> post-zygotic isolation. Aggression between <strong>male</strong>s can lead directly toselection aga<strong>in</strong>st hybrids if hybrids receive a disproportionate amount <strong>of</strong> attack by <strong>male</strong>s <strong>of</strong> both parental species.Population divergence <strong>in</strong> traits used <strong>in</strong> <strong>male</strong> <strong>competition</strong> can also lead to selection aga<strong>in</strong>st hybrids as side effects. Future<strong>in</strong>vestigations <strong>of</strong> the relative importance <strong>of</strong> divergence <strong>in</strong> traits used <strong>in</strong> <strong>male</strong> <strong>contest</strong> compared to population divergence<strong>in</strong> other traits are needed if we want to understand the <strong>role</strong> <strong>of</strong> <strong>male</strong> <strong>competition</strong> <strong>in</strong> the <strong>speciation</strong> process.6 Conclusions and future prospects<strong>The</strong> whole <strong>speciation</strong> process, from the <strong>in</strong>itial start <strong>of</strong> divergence between populations to the evolution <strong>of</strong> competereproductive isolation between them, <strong>of</strong>ten takes a very long time. Researchers <strong>in</strong>terested <strong>in</strong> the mechanisms driv<strong>in</strong>g theprocess <strong>of</strong> <strong>speciation</strong> are therefore, based on the current stage <strong>of</strong> population divergence that they study, faced with theproblem <strong>of</strong> either hav<strong>in</strong>g to predict the future and/or reconstruct history.<strong>The</strong> <strong>role</strong> <strong>of</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>in</strong> promot<strong>in</strong>g and ma<strong>in</strong>ta<strong>in</strong><strong>in</strong>g diversity rema<strong>in</strong>s a little explored scientific area.Most studies carried out so far have focused on the effects <strong>of</strong> <strong>male</strong> aggression towards similar competitors at earlystages <strong>of</strong> population divergence (Seehausen and Schluter, 2004; Dijkstra et al., 2006; Young et al., 2009). <strong>The</strong> majorconclusions are that aggression towards similar competitors may facilitate both the establishment <strong>of</strong> new color morphsand the ma<strong>in</strong>tenance <strong>of</strong> polymorphism. However, there are also possible ways by which such negative frequencydependentselection may prevent population divergence. It would, for example, promote successful immigration <strong>of</strong><strong>male</strong>s between populations that have started to diverge <strong>in</strong> allopatry, s<strong>in</strong>ce rare phenotypes are favored. <strong>The</strong>homogeniz<strong>in</strong>g effects <strong>of</strong> gene flow could then prevent population differentiation (Mayr, 1963; Felsenste<strong>in</strong>, 1981; Coyneand Orr, 2004). Future studies on the effects <strong>of</strong> <strong>male</strong> aggression towards similar competitors at different stages <strong>of</strong>population divergence and under different geographical conditions are needed.As populations adapt to their environment, complexes <strong>of</strong> traits may change non-<strong>in</strong>dependently <strong>of</strong> each other, eitherdue to genetic correlations between these traits or because they have synergistic effects on fitness. <strong>The</strong> more traits thatdiverge between populations, the more likely is the build up <strong>of</strong> reproductive isolation between them (Nosil et al., 2009).One question that we raised here was how adaptation to different ecological environments co<strong>in</strong>cides with divergence <strong>in</strong>traits used <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> and vice versa. S<strong>in</strong>ce aggressive abilities are likely to trade-<strong>of</strong>f aga<strong>in</strong>st a number<strong>of</strong> other crucial abilities such as parasite resistance, protection aga<strong>in</strong>st predators and general stress tolerance, we expectwhole suites <strong>of</strong> traits to diverge non-randomly across populations experienc<strong>in</strong>g different ecological environments. Wewould like to encourage more future studies exam<strong>in</strong><strong>in</strong>g how divergence <strong>in</strong> traits used <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong>, ascompared to other traits, <strong>in</strong>fluences the build up <strong>of</strong> reproductive isolation. However, the key to understand<strong>in</strong>g the <strong>role</strong> <strong>of</strong><strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>in</strong> <strong>speciation</strong> probably lies <strong>in</strong> the exam<strong>in</strong>ation <strong>of</strong> how the different selection pressures(<strong>in</strong>clud<strong>in</strong>g natural selection and the two ma<strong>in</strong> mechanisms <strong>of</strong> sexual selection) <strong>in</strong>teract dur<strong>in</strong>g the different stages <strong>of</strong>population divergence.When populations diverge <strong>in</strong> sympatry, harmful <strong>in</strong>terspecific <strong>in</strong>teractions such as <strong>competition</strong> <strong>over</strong> <strong>mates</strong> and/orresources can be reduced by divergence <strong>in</strong> resource-use or <strong>in</strong> reproductive phenotypes, i.e. character displacement12
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>(Brown and Wilson, 1956). <strong>The</strong> importance <strong>of</strong> <strong>in</strong>terspecific aggression <strong>in</strong> driv<strong>in</strong>g character displacement is ga<strong>in</strong><strong>in</strong>g<strong>in</strong>creas<strong>in</strong>g support (Seehausen and Schluter, 2004; Tynkkynen et al., 2004; 2005; Grether et al., 2009; Anderson andGrether, 2010, Vall<strong>in</strong> et al., 2011b). One particularly <strong>in</strong>terest<strong>in</strong>g open research question is how character displacementthrough <strong>in</strong>terspecific aggression <strong>in</strong>fluences re<strong>in</strong>forcement, i.e. selection aga<strong>in</strong>st hybridization driv<strong>in</strong>g divergence <strong>in</strong>mat<strong>in</strong>g traits.Male <strong>contest</strong> <strong>competition</strong> <strong>over</strong> fe<strong>male</strong>s (or <strong>over</strong> resources needed to attract fe<strong>male</strong>s) can facilitate <strong>speciation</strong> bycaus<strong>in</strong>g direct selection aga<strong>in</strong>st hybrids or by lead<strong>in</strong>g to fast population divergence, which <strong>in</strong> turn may lead toreproductive isolation as a side effect through a number <strong>of</strong> different pathways. However, we would also like to stressthat <strong>male</strong> aggression under certa<strong>in</strong> conditions may also limit assortative mat<strong>in</strong>g and/or population divergence. In thisreview, we have outl<strong>in</strong>ed some examples <strong>of</strong> how sexual selection through <strong>male</strong> <strong>competition</strong> can <strong>in</strong>fluence <strong>speciation</strong>.<strong>The</strong>re is plenty <strong>of</strong> empirical evidence show<strong>in</strong>g that <strong>male</strong> <strong>contest</strong> <strong>competition</strong> can cause rapid evolutionary changes, buthow these evolutionary changes <strong>in</strong>fluence <strong>speciation</strong> rema<strong>in</strong>s surpris<strong>in</strong>gly unexplored.Acknowledgements We would like to thank M. Servedio, A. Rice and three anonymous reviewers for constructive comments. Weare supported by the Swedish Research Council, the Royal Swedish Academy <strong>of</strong> Sciences and the European Research Foundation bygrants to AQ.ReferencesAlatalo RV, Gustafsson L, Lundberg A, 1994. Male coloration and species recognition <strong>in</strong> sympatric flycatchers. Proc. R. Soc. Lond. B. 256:113–118.Alcock J, 1987. <strong>The</strong> effects <strong>of</strong> experimental manipulation <strong>of</strong> resources on the behavior <strong>of</strong> two calopterygid damselflies that exhibit resource–defense polygyny. Can. J. Zool.-Rev. Canad. De Zool. 65: 2475–2482.Anderson CN, Grether GF, 2010. Interspecific aggression and character displacement <strong>of</strong> competitor recognition <strong>in</strong> Hetaer<strong>in</strong>a damselflies.Proc. R. Soc. Lond. B. 277: 549–555.Andersson M, 1994. Sexual Selection. Pr<strong>in</strong>ceton, NJ: Pr<strong>in</strong>ceton University Press.Arnkvist G, Rowe L, 2002. Antagonistic coevolution between the sexes <strong>in</strong> a group <strong>of</strong> <strong>in</strong>sects. Nature 415: 787–789.Arnqvist G, Rowe L, 2005. Sexual Conflict. Pr<strong>in</strong>ceton, NJ: Pr<strong>in</strong>ceton University Press.Bateson PPG, 1966. <strong>The</strong> characteristics and context <strong>of</strong> impr<strong>in</strong>t<strong>in</strong>g. Biol. Rev. 41: 177–220.Berenbaum M, 1983. Coumar<strong>in</strong>s and caterpillars: A case for coevolution. Evolution 37: 163–179.Berglund A, Bisazza A, Pilastro A, 1996. Armaments and ornaments: An evolutionary explanation <strong>of</strong> traits <strong>of</strong> dual utility. Biol. J. L<strong>in</strong>n. Soc.58: 385–399.Birkhead TR, Mart<strong>in</strong>ez JG, Burke T, Froman DP, 1999. Sperm mobility determ<strong>in</strong>es the outcome <strong>of</strong> sperm <strong>competition</strong> <strong>in</strong> the domestic fowl.Proc. R. Soc. Lond. B. 266: 1759–1764.Bouton N, Seehausen O, van Alphen JJM, 1997. Resource partition<strong>in</strong>g among rock dwell<strong>in</strong>g haplochrom<strong>in</strong>es (Pisces: Cichlidae) from LakeVictoria. Ecol. Freshw. Fish 6: 225–240.Brazil-Boast J, van Rooij E, Pryke SR, Griffith SC, 2011. Interference from Long-tailed f<strong>in</strong>ches constra<strong>in</strong>s reproduction <strong>in</strong> the endangeredgoldian f<strong>in</strong>ch. J. Animal. Ecol. 80: 39–48.Briffa M, Sneddon LU, 2007. Physiological constra<strong>in</strong>ts on <strong>contest</strong> behaviour. Func. Ecol. 21: 627–637.Brown WL, Wilson EO, 1956. Character displacement. Syst. Zool. 5: 49–64.Butl<strong>in</strong> RK, 1995. Re<strong>in</strong>forcement: An idea evolv<strong>in</strong>g. Trends. Ecol. Evol. 10: 432–434.Candol<strong>in</strong> U, 2004. Oppos<strong>in</strong>g selection on a sexually dimorphic trait through fe<strong>male</strong> choice and <strong>male</strong> <strong>competition</strong> <strong>in</strong> a water boatman.Evolution 58; 1861–1864.13
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>Chellappa S, Hunt<strong>in</strong>gford FA, 1989. Depletion <strong>of</strong> energy reserves dur<strong>in</strong>g reproductive aggression <strong>in</strong> <strong>male</strong> three-sp<strong>in</strong>ed sticklebackGasterosteus aculeatus. J. Fish Biol. 35: 315–316.Clayton NS, 1993. Song, sex and sensitive phases <strong>in</strong> the behavioural development <strong>of</strong> birds. Trends. Ecol. Evol. 4: 82–84.Cooley JR, 2007. Decod<strong>in</strong>g asymmetries <strong>in</strong> reproductive character displacement. Proc. Acad. Nat. Sci. Phil. 156: 89–96.Coyne JA, Orr HA, 1989. Patterns <strong>of</strong> <strong>speciation</strong> <strong>in</strong> Drosophila. Evolution 43: 362–381.Crothers L, Ger<strong>in</strong>g E, Cumm<strong>in</strong>gs M, 2011. Aposematic signal variation predicts <strong>male</strong>-<strong>male</strong> <strong>in</strong>teractions <strong>in</strong> a polymorphic poison prog.Evolution 65: 599–605.Dieckmann U, Doebeli M, 1999. On the orig<strong>in</strong> <strong>of</strong> species by sympatric <strong>speciation</strong>. Nature 400: 354–357.Dieckmann U, Doebeli M, Metz JA, Tautz D, 2004. Adaptive <strong>speciation</strong>. Cambridge: Cambridge University Press.Dijkstra PD, Seehausen O, Pierotti, MER, Groothuis TGG, 2007. Male-<strong>male</strong> <strong>competition</strong> and <strong>speciation</strong>: Aggression bias towards differentlycoloured rivals varies between stages <strong>of</strong> <strong>speciation</strong> <strong>in</strong> a Lake Victoria cichlid species complex. J. Evol. Biol. 20: 496–502.Dijkstra PD, Groothuis TGG, 2011. Male-<strong>male</strong> <strong>competition</strong> as a force <strong>in</strong> evolutionary diversification: evidence <strong>in</strong> haplochrom<strong>in</strong>e cichlid fish.Int. J. Evol. Biol. 2011, Article ID 689254.Dobzhansky T, 1935. A critique <strong>of</strong> the species concept <strong>in</strong> biology. Phil. <strong>of</strong> Science 2: 344–355.Dobzhansky T, 1940. Speciation as a stage <strong>in</strong> evolutionary divergence. Am. Nat. 74: 312–321.Dom<strong>in</strong>ey W, 1984. Effects <strong>of</strong> sexual selection and life history on <strong>speciation</strong>: Species flocks <strong>in</strong> African cichlids and Hawaiian Drosphila. In:Echelle AA, Kornfield I ed. Evolution <strong>of</strong> Fish Species Flocks. Orono, ME: University <strong>of</strong> Ma<strong>in</strong>e Press, 231–249.van Doorn GS, Dieckmann U, Weiss<strong>in</strong>g FJ, 2004. Sympatric <strong>speciation</strong> by sexual selection, a critical re-evaluation. Am. Nat. 163: 709–725.Van Doorn GS, Edelaar P, Weiss<strong>in</strong>g FJ, 2009. On the orig<strong>in</strong> <strong>of</strong> species by natural and sexual selection. Science 326: 1704–1707.Drost R, 1936. Über das Brutkleid männlicher Trauerfliegenfänger Muscicapa hypoleuca. Vogelzug 6: 179–186.Dunn M, Copelston M, Workman L, 2004. Trade-<strong>of</strong>fs and seasonal variation <strong>in</strong> territorial defense and predator evasion <strong>in</strong> the EuropeanRob<strong>in</strong> Erithacus rubecula. Ibis 146: 77–84.Edwards SV, K<strong>in</strong>gan SB, Calk<strong>in</strong>s JD, Balakrishnan CN, Jenn<strong>in</strong>gs WB et al., 2005. Speciation <strong>in</strong> birds: Genes, geography, and sexualselection. Proc. Natl. Acad. Sci. USA. 102: 6550–6557.Elmer KR, Kusche H, Lehtonen TK, Meyer A, 2010. Local variation and parallel evolution: morphological and genetic diversity across aspecies complex <strong>of</strong> neotropical crater lake cichlid fishes. Philos. T. R. Soc. B. 365: 1763–1782.Fabricius E, 1991. Interspecific mate choice follow<strong>in</strong>g cross-foster<strong>in</strong>g <strong>in</strong> a mixed colony <strong>of</strong> greylag geese Anser anser and canada geeseBranta Canadensis: A study on development and persistence <strong>of</strong> species preferences. Ethology 88: 287–296.Farr JA, 1977. Male rarity or novelty, fe<strong>male</strong> choice behaviour, and sexual selection <strong>in</strong> the guppy Poecilia reticulata. Evolution 31: 162–168.Feder JL, Opp SB, Wlazlo B, Reynolds K, Go W et al., 1994. Host fidelity is an effective premat<strong>in</strong>g barrier between sympatric races <strong>of</strong> theapple maggot fly. Proc. Natl. Acad. Sci. USA. 91: 7990–7994.Felsenste<strong>in</strong> J, 1981. Skepticism towards Santa Rosalia, or why are there so few k<strong>in</strong>ds <strong>of</strong> animals? Evolution 35: 124–138.Gavrilets S 2000. Rapid evolution <strong>of</strong> reproductive barriers driven by sexual conflict. Nature 403: 886–889.Gavrilets S, 2004. Fitness landscapes and the orig<strong>in</strong> <strong>of</strong> species. Pr<strong>in</strong>ceton, NJ: Pr<strong>in</strong>ceton University Press.Grant BR, Grant PR, 1993. Evolution <strong>of</strong> Darw<strong>in</strong> f<strong>in</strong>ches caused by a rare climatic event. Proc. R. Soc. Lond. B. 252: 253–253.Grant PR, Grant BR, 1997. Genetics and the orig<strong>in</strong> <strong>of</strong> bird species. Proc. Natl. Acad. Sci. USA. 94: 7768–7775.Grear DA, Perk<strong>in</strong>s SE, Hudson PJ, 2009. Does elevated testosterone result <strong>in</strong> <strong>in</strong>creased exposure and transmission <strong>of</strong> parasites? Ecol. Letters12:528–537.Grether GF, Los<strong>in</strong> N, Anderson CN, Okamoto K, 2009. <strong>The</strong> <strong>role</strong> <strong>of</strong> <strong>in</strong>terspecific <strong>in</strong>terference <strong>competition</strong> <strong>in</strong> character displacement and theevolution <strong>of</strong> competitor recognition. Biol. Rev. 84: 617–635.Gross MR, 1996. Alternative reproductive strategies and tactics: diversity with<strong>in</strong> sexes. Trends Ecol. Evol. 11: 92–98.Grön<strong>in</strong>g J, Hochkirch A, 2008. Reproductive <strong>in</strong>terference between animal species. Q. Rev. Biol. 83: 257–282.Gustafsson L, Pärt T, 1991. Interspecific relations between collared and pied flycatchers. Proc. Int. Orn. Con. 20: 1425–1431.Haldane JBS, 1922. Sex–ratio and unisexual sterility <strong>in</strong> hybrid animals. J. Gen. 12:101–109.Harris MP, 1970. Abnormal migration and hybridization <strong>of</strong> Larus argentatus and L. fuscus after <strong>in</strong>terspecies foster<strong>in</strong>g experiments. Ibis 112:488–498.Hendry AP, Day T, 2005. Population structure attributable to reproductive time: Isolation by time and adaptation by time. Mol. Ecol. 14:901–916.Higashi M, Takimoto G, Yamamura N, 1999. Sympatric <strong>speciation</strong> by sexual selection. Nature 402: 523–526.14
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>Hoekstra HE, Hirschmann RJ, Bundey RA, Insel PA, Crossland JP, 2006. A s<strong>in</strong>gle am<strong>in</strong>o acid mutation contributes to adaptive beach mousecolor pattern. Science 313: 101–104.H<strong>of</strong>reiter M, Schöneberg T, 2010. <strong>The</strong> genetic and evolutionary basis <strong>of</strong> colour variation <strong>in</strong> vertebrates. Cell. Mol. Life. Sci. 67: 2591–2603.Hosk<strong>in</strong> CJ, Higgie M, 2010. Speciation via species <strong>in</strong>teractions: the divergence <strong>of</strong> mat<strong>in</strong>g traits with<strong>in</strong> species. Ecol. Lett. 13:409–420.Hosk<strong>in</strong> CJ, Higgie M, McDonald KR, Moritz C, 2005. Re<strong>in</strong>forcement drives rapid allopatric <strong>speciation</strong>. Nature 437:1353–1356.Hostert EE, 1997. Re<strong>in</strong>forcement: A new perspective on an old contr<strong>over</strong>sy. Evolution 51: 697–702.Howard DJ, 1993. Re<strong>in</strong>forcement: Orig<strong>in</strong>, dynamics, and fate <strong>of</strong> an evolutionary hypothesis. In: Harrison RG ed. Hybrid Zones and theEvolutionary Process. New York, New York, USA: Oxford University Press, 46–69.Huhta E, Siikamäki P, Jokimäki J, 1997. Small scale geographical variation <strong>in</strong> plumage colour <strong>of</strong> pied flycatcher <strong>male</strong>s. J. Avian. Biol. 28:92–94.Hunt J, Simmons LW, 2000. Maternal and paternal effects on <strong>of</strong>fspr<strong>in</strong>g phenotype <strong>in</strong> the dung beetle Onthophagus taurus. Evolution 54:936–941.Hunt J, Breuker CJ, Sadowski JA, Moore AJ, 2009. Male-<strong>male</strong> <strong>competition</strong>, fe<strong>male</strong> mate choice and their <strong>in</strong>teraction: Determ<strong>in</strong><strong>in</strong>g totalsexual selection. J. Evol. Biol. 22: 13–26.Hunt<strong>in</strong>gford FA, Turner AK, 1987. Animal conflict. London: Chapman and Hall.Hurtado-Gonzales JL, Uy JAC, 2009. Alternative mat<strong>in</strong>g strategies may favour the persistence <strong>of</strong> a genetically based colour polymorphism<strong>in</strong> a pentamorphic fish. Anim. Behav. 77: 1187–1194.Jakobsson S, Brick O, Kullberg C, 1995. Escalated fight<strong>in</strong>g behaviour <strong>in</strong>curs <strong>in</strong>creased predation risk. Anim. Behav. 49: 235–239.Irw<strong>in</strong> DE, Price T, 1999. Sexual impr<strong>in</strong>t<strong>in</strong>g, learn<strong>in</strong>g and <strong>speciation</strong>. Heredity 82: 347–354.Kimura K, Chiba S, 2010. Interspecific <strong>in</strong>terference <strong>competition</strong> alters habitat use patterns <strong>in</strong> two species <strong>of</strong> land snails. Evol. Ecol. 24: 815–825.Kirkpatrick M, 1987. Sexual selection by fe<strong>male</strong> choice <strong>in</strong> polygynous animals. Ann. Rev. Ecol. Syst. 18: 43–70.Kirkpatrick M, Ravigné V, 2002. Speciation by natural and sexual selection: models and experiments. Am. Nat. 159: 22–35.Kirkpatrick M, Nuismer SL, 2004. Sexual selection can constra<strong>in</strong> sympatric <strong>speciation</strong>. Proc. R. Soc. Lond. B 271:687–693.Knight ME, Turner GF, 2004. Laboratory mat<strong>in</strong>g trials <strong>in</strong>dicate <strong>in</strong>cipient <strong>speciation</strong> by sexual selection among populations <strong>of</strong> the cichlid fishPseudotropheus zebra from Lake Malawi. Proc. Roy. Soc. Lond. B 271:675–680.Kozak GM, Head ML, Boighman JW, 2011. Sexual impr<strong>in</strong>t<strong>in</strong>g on ecologically divergent traits leads to sexual isolation <strong>in</strong> sticklebacks. Proc.Roy. Soc. Lond. B 278:2604–2610.Kraaijeveld K, Kraaijeveld-Smit FJL, Maan ME, 2011. Sexual selection and <strong>speciation</strong>: <strong>The</strong> comparative evidence revisited. Biol. Rev.86: 367–377Král M, Järvi T, Biclk V, 1988. Inter-specific aggression between the collared flycatcher and the pied flycatcher: <strong>The</strong> selective agent for theevolution <strong>of</strong> light colored <strong>male</strong> pied flycatcher populations. Ornis. Scand. 19: 287–289.Labonne J, Hendry AP, 2010. Natural and sexual selection giveth and taketh away reproductive barriers: Models <strong>of</strong> population divergence <strong>in</strong>guppies. Am. Nat. 176: 26–39.Lande R, 1981. Models <strong>of</strong> <strong>speciation</strong> by sexual selection on polygenic traits. Proc. Natl. Acad. Sci. USA. 78: 3721–3725.Lank DB, Smith CM, Hanotte O, Burke TA, Cooke F, 1995. Genetic polymorphism for alternative mat<strong>in</strong>g behavior <strong>in</strong> lekk<strong>in</strong>g <strong>male</strong> ruffPhilomachus pugnax. Nature 378: 59–62.Maan ME, Cumm<strong>in</strong>gs ME, 2008. Fe<strong>male</strong> preferences for aposematic signal components <strong>in</strong> a polymorphic poison frog. Evolution 62: 2334–2345.Maan ME, Seehausen O, 2011. Ecology, sexual selection and <strong>speciation</strong>. Ecol. Lett. 14:591–602.Mallet J, 1989. <strong>The</strong> genetics <strong>of</strong> warn<strong>in</strong>g colour <strong>in</strong> Peruvian hybrid zones <strong>of</strong> Heliconius erato and H. melpomene. Proc. R. Soc. Lond. B. 236:163–185.Manceau M, Dom<strong>in</strong>gues VS, Mallar<strong>in</strong>o R, Hoekstra HE, 2011. <strong>The</strong> developmental <strong>role</strong> <strong>of</strong> agouti <strong>in</strong> color pattern evolution. Science 331:1062–1065.Mart<strong>in</strong>ez-Sanchis S, Arnedo MT, Salvador A, Moya-Albiol L, Gonzales-Bono E, 2003. Effects <strong>of</strong> chronic adm<strong>in</strong>istration with high doses <strong>of</strong>testosterone propionate on behavioral and physiological parameters <strong>in</strong> mice with differ<strong>in</strong>g basal aggressiveness. Aggress. Behav. 29:173–189.Mayr E, 1963. Animal species and Evolution. Cambridge: Belknap Press.Mendelson TC, 2003. Sexual isolation evolves faster than hybrid <strong>in</strong>viability <strong>in</strong> a diverse and sexually dimorphic genus <strong>of</strong> fish (Percidae:Etheostoma). Evolution 57:317–327.15
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>Mikami OK, Kohda M, Kawata M, 2004. A new hypothesis for species co-existence, <strong>male</strong>-<strong>male</strong> repulsion promotes co-existence <strong>of</strong>compet<strong>in</strong>g species. Pop. Ecol. 46: 213–217.Milligan BG, 1985. Evolutionary divergence and character displacement <strong>in</strong> two phenotypically-variable, compet<strong>in</strong>g species. Evolution 39:1207–1222.Moczek AP, Emlen DJ, 1999. Proximate determ<strong>in</strong>ation <strong>of</strong> <strong>male</strong> horn dimorphism <strong>in</strong> the beetle Onthophagus taurus (Coleoptera:Scarabaeidae). J. Evol. Biol. 12:27–37.Moczek AP, Emlen DJ, 2000. Male horn dimorphism <strong>in</strong> the scarab beetle Onthophagus taurus: Do alternative reproductive tactics favoralternative phenotypes? Anim. Behav, 59: 459–466.Mougeot F, Irv<strong>in</strong>e JR, Seivwright L, Redpath SM, Piertney S, 2004. Testosterone, immunocompetence, and honest sexual signal<strong>in</strong>g <strong>in</strong> <strong>male</strong>red grouse. Behav. Ecol. 15: 930–937.Mougeot F, Evans SA, Redpath SM, 2005. Interactions between population processes <strong>in</strong> a cyclic species: Parasites reduce autumn territorialbehaviour <strong>of</strong> <strong>male</strong> red grouse. Oecologia 144: 289–298.Muller HJ, 1940. Bear<strong>in</strong>g <strong>in</strong> the Drosophila work on systematics. In: Huxley JS ed. <strong>The</strong> New Systematics. Oxford: Clarenton Press, 185–268.Nagl S, Tichy H, Mayer WE, Takezaki N, Takahata N et al., 2000. <strong>The</strong> orig<strong>in</strong> and age <strong>of</strong> haplochrom<strong>in</strong>e fishes <strong>in</strong> Lake Victoria, East Africa.Proc. R. Soc. Lond. B. 267: 1049–1061.Nosil P, Crespi BJ, Sandoval CP, 2003. Reproductive isolation driven by the comb<strong>in</strong>ed effects <strong>of</strong> ecological adaptation and re<strong>in</strong>forcement.Proc. R. Soc. Lond. B. 207: 1911–1918.Nosil P, Harmon LJ, Seehausen O, 2009. Ecological explanations for (<strong>in</strong>complete) <strong>speciation</strong>. Trends. Ecol. Evol. 24:145–156.Noor MA, 1995. Speciation driven by natural selection <strong>in</strong> Drosophila. Nature 375: 674–675.Noor MAF, 1999. Re<strong>in</strong>forcement and other consequences <strong>of</strong> sympatry. Heredity 83: 503–508.Ortiz-Barrientos D, Grealy A, Nosil, P. 2009. <strong>The</strong> genetics and ecology <strong>of</strong> re<strong>in</strong>forcement: Implications for the evolution <strong>of</strong> prezygoticisolation <strong>in</strong> sympatry and beyond. Ann. Ny. Acad. Sci. 1168: 156–182.Panhuis TM, Butl<strong>in</strong> R, Zuk M, Tregenza T, 2001. Sexual selection and <strong>speciation</strong>. Trends Ecol. Evol. 16: 364–371.Parker GA, 1979. Sexual selection and sexual conflict. In: Blum MS, Blum NA ed. Sexual Selection and Reproductive Competition <strong>in</strong>Insects. New York: Academic, 23–166.Pearce D, Pryke SR, Griffith SC, 2011. Interspecific aggression for nest sites: Model experiments with long-tailed f<strong>in</strong>ches Poephillaacuticauda and endangered gouldian f<strong>in</strong>ches Erythrura gouldiae. Auk 128: 497–505.Pfennig KS, Pfennig DW, 2009. Character displacement: Ecological and reproductive responses to a common evolutionary problem. Q. Rev.Biol. 84: 253–276.Polak M, Starmer WT, 1998. Parasite-<strong>in</strong>duced risk <strong>of</strong> mortality elevates reproductive effort <strong>in</strong> <strong>male</strong> Drosophila. Proc. R. Soc. Lond. B. 265:2197–2201.Price T, 1998. Sexual selection and natural selection <strong>in</strong> bird <strong>speciation</strong>. Philos. T. R. Soc. B. 353: 251–260.Price TD, 2002. Domesticated birds as a model for the genetics <strong>of</strong> <strong>speciation</strong> by sexual selection. Genetica 116: 311–327.Price TD, Bouvier MM, 2002. <strong>The</strong> evolution <strong>of</strong> F1 postzygotic <strong>in</strong>compatibilities <strong>in</strong> birds. Evolution 56:2083–2089.Pröhl H, 2005. Territorial behavior <strong>in</strong> dendrobatid frogs. J. Herpet. 39: 354–365.Qvarnström A, Forsgren E, 1998. Should fe<strong>male</strong>s prefer dom<strong>in</strong>ant <strong>male</strong>s? Trends Ecol. Evol. 13: 498–501.Qvarnström A, Blomgren V, Wiley C, Sved<strong>in</strong> N, 2004. Fe<strong>male</strong> collared flycatchers learn to prefer <strong>male</strong>s with an artificial novel ornament.Behav. Ecol. 15: 543–548.Qvarnström A, Brommer JE, Gustafsson L, 2006. Test<strong>in</strong>g the genetics underly<strong>in</strong>g the co-evolution <strong>of</strong> mate choice and ornament <strong>in</strong> the wild.Nature 441: 84–86.Qvarnström A, Wiley C, Sved<strong>in</strong> N, Vall<strong>in</strong> N, 2009. Life-history divergence facilitates regional coexistence <strong>of</strong> compet<strong>in</strong>g Ficedulaflycatchers. Ecology 90: 1948–1957.Reynolds RG, Fitzpatrick BM, 2007. Assortative mat<strong>in</strong>g <strong>in</strong> poison-dart frogs based on an ecologically important trait. Evolution 61: 2253–2259.Rice WR, Chippendale AK, 2001. Sexual recomb<strong>in</strong>ation and the power <strong>of</strong> natural selection. Science 294: 555–559.Rice AM, Rudh A, Ellegren H, Qvarnström A, 2011. A guide to the genomics <strong>of</strong> ecological <strong>speciation</strong> <strong>in</strong> natural animal populations. Ecol.Lett. 14: 9–18.Ritchie MG, 2007. Sexual selection and <strong>speciation</strong>. Annu. Rev. Ecol. Evol. S. 38: 79–102.Roskaft E, Järvi T, 1992. Interspecific <strong>competition</strong> and the evolution <strong>of</strong> plumage-color variation <strong>in</strong> three closely related old-world flycatchersFicedula spp. J. Zool. 228: 521–532.16
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>Rudh A, Rogell B, Håstad O, Qvarnström A, 2011. Rapid population divergence l<strong>in</strong>ked with co-variation between coloration and sexualdisplay <strong>in</strong> Strawberry poison frogs. Evolution 65: 1271–1282.Rundle HD, Nosil P, 2005. Ecological <strong>speciation</strong>. Ecol. Lett. 8: 336–352.Saporito RA, Zuercher R, Roberts M, Gerow KG, Donnelly MA, 2007. Experimental evidence for aposematism <strong>in</strong> the dendrobatid poisonfrog Oophaga pumilio. Copeia 4: 1006–1011.Sætre G-P, Král M, Bičík V, 1993. Experimental evidence for <strong>in</strong>terspecific fe<strong>male</strong> mimcry <strong>in</strong> sympatric Ficedula flycatchers. Evolution 47:939–945.Sætre G-P, Moum T, Bures, T, Král M, Adamjan M, Moreno J, 1997. A sexually selected character displacement <strong>in</strong> flycatchers re<strong>in</strong>forcespremat<strong>in</strong>g isolation. Nature 387: 589–592.Schluter D, 2000. <strong>The</strong> Ecology <strong>of</strong> Adaptive Radiation. Oxford: Oxford University Press.Schluter D, Price T, 1993. Honesty, perception and population divergence <strong>in</strong> sexually selected traits. Proc. R. Soc. Lond. B. 253: 117–122.Seehausen O, Bouton N, 1997. Microdistribution and fluctuations <strong>in</strong> niche <strong>over</strong>lap <strong>in</strong> a rocky shore cichlid community <strong>in</strong> Lake Victoria. Ecol.Freshw. Fish 6: 161–173.Seehausen O, Schluter D, 2004. Male-<strong>male</strong> <strong>competition</strong> and nuptial-colour displacement as a diversify<strong>in</strong>g force <strong>in</strong> Lake Victoria cichlidfishes. Proc. R. Soc. Lond. B. 271: 1345–1353.Seehausen O, van Alphen JJM, 1999. Can sympatric <strong>speciation</strong> by disruptive sexual selection expla<strong>in</strong> rapid evolution <strong>of</strong> cichlid diversity <strong>in</strong>Lake Victoria? Ecol. Lett. 2: 262–271.Servedio MR, 2009. <strong>The</strong> <strong>role</strong> <strong>of</strong> l<strong>in</strong>kage disequilibrium <strong>in</strong> the evolution <strong>of</strong> premat<strong>in</strong>g isolation. Heredity 102: 51–56.Servedio MR, van Doorn GS, Kopp M, Frame AM, Nosil P, 2011. Magic traits <strong>in</strong> <strong>speciation</strong>: ‘magic’ but not rare? Trends. Ecol. Evol. 26:389–397.Shuster SM, Wade MJ, 1991. Equal mat<strong>in</strong>g success among <strong>male</strong> reproductive strategies <strong>in</strong> a mar<strong>in</strong>e isopod. Nature 350: 608–610.Siddiqi A, Cron<strong>in</strong> TW, Loew ER, Vorobyev M, Summers K, 2004. Interspecific and <strong>in</strong>traspecific views <strong>of</strong> color signals <strong>in</strong> the strawberrypoison frog Dendrobates pumilio. J. Exp. Biol. 207: 2471–2485.S<strong>in</strong>ervo A, Lively CM, 1996, <strong>The</strong> rock-paper-scissors game and the evolution <strong>of</strong> alternative <strong>male</strong> strategies. Nature 380: 240–243.Summers K, Symula R, Clough M, Cron<strong>in</strong> T, 1999. Visual mate choice <strong>in</strong> poison frogs. Proc. R. Soc. Lond. B. 266: 2141–2145.Summers K, Clough ME, 2001. <strong>The</strong> evolution <strong>of</strong> coloration and toxicity <strong>in</strong> the poison frog family (Dendrobatidae). Proc. Natl. Acad. Sci.USA. 98: 6227–6232.Summers K, Cron<strong>in</strong> TW, Kennedy T, 2003. Variation <strong>in</strong> spectral reflectance among populations <strong>of</strong> Dendrobates pumilio, the strawberrypoison frog, <strong>in</strong> the Bocas del Toro Archipelago, Panama. J. Biogeo. 30: 35–53.Svensson EI, Gosden TP, 2007. Contemporary evolution <strong>of</strong> secondary sexual traits <strong>in</strong> the wild. Func. Ecol. 21: 422–433.Svensson EI, Eroukhman<strong>of</strong>f F, Karlsson K, Runemark A, Brod<strong>in</strong> A, 2010.<strong>The</strong> <strong>role</strong> <strong>of</strong> learn<strong>in</strong>g <strong>in</strong> population divergence <strong>of</strong> mate preference.Evolution 64: 1301–3113.Théron A, Combes C, 1995. Asynchrony <strong>of</strong> <strong>in</strong>fection tim<strong>in</strong>g, habitat preference, and sympatric <strong>speciation</strong> <strong>of</strong> Schistosome parasites.Evolution 49: 372–375.Trivers R, 1972. Parental <strong>in</strong>vestment and sexual selection. In: Champell B ed. Sexual Selection and the Descent <strong>of</strong> Man. London: He<strong>in</strong>emann,136–179.Tuttle EM, 2003. Alternative reproductive strategies <strong>in</strong> the white-throated sparrow: behavioral and genetic evidence. Behav. Ecol. 14: 425–432.Tynkkynen K, Rantala MJ, Suhonen J, 2004. Interspecific aggression and character displacement <strong>in</strong> the damselfly Calopteryx splendens. J.Evol. Biol. 17: 759–767.Tynkkynen K, Kotiaho JS, Luojumaki M, Suhonen J, 2005. Interspecific aggression causes negative selection on sexual characters.Evolution 59: 1838–1843.Tynkkynen K, Kotiaho JS, Luojumaki M, Suhonen J, 2006. Interspecific territoriality <strong>in</strong> Calopteryx damselflies: <strong>The</strong> <strong>role</strong> <strong>of</strong> secondarysexual characters. Anim. Behav. 71: 299–306.Vall<strong>in</strong> N, Qvarnström A, 2011 . Learn<strong>in</strong>g the hard way: Impr<strong>in</strong>t<strong>in</strong>g can enhance enforced shifts <strong>in</strong> habitat choice. Int. J. Ecol., <strong>in</strong> press.Vall<strong>in</strong> N, Rice AM, Artsen H, Kulma K, Qvarnström A, 2011a. Comb<strong>in</strong>ed effects <strong>of</strong> <strong>in</strong>terspecific <strong>competition</strong> and hybridization impedelocal coexistence <strong>of</strong> Ficedula flycatchers. Evol. Ecol., <strong>in</strong> pressVall<strong>in</strong> N, Rice AM, Bailey RI, Husby A, Qvarnström A, 2011b. Positive feedback between ecological and reproductive characterdisplacement <strong>in</strong> a young avian hybrid zone, Evolution, <strong>in</strong> press.17
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>Wang IJ, Shaffer HB, 2008. Rapid color evolution <strong>in</strong> an aposematic species: A phylogenetic analysis <strong>of</strong> color variation <strong>in</strong> the strik<strong>in</strong>glypolymorphic strawberry poison-dart frog. Evolution 62: 2742–2759.West-Eberhard MJ, 1983. Sexual selection, social <strong>competition</strong>, and <strong>speciation</strong>. Q. Rev. Biol. 58: 155–183.Wirtz P, 1999. Mother species - father species: Unidirectional hybridization <strong>in</strong> animals with fe<strong>male</strong> choice. Anim. Behav. 58: 1–12.Wong BBM, Candol<strong>in</strong> U, 2005. How is fe<strong>male</strong> mate choice affected by <strong>male</strong> <strong>competition</strong>? Biol. Rev. 80: 559–571.Wu C-I, Davies AW, 1993. Evolution <strong>of</strong> post-mat<strong>in</strong>g reproductive isolation: <strong>The</strong> composite nature <strong>of</strong> Haldane’s rule and its genetic basis.Am. Nat. 142:187–212.Wu C-I. Johnson NA, Palopolo MF, 1996. Haldane’s rule and its legacy: Why are there so many sterile <strong>male</strong>s? Trends. Ecol. Evol. 11: 281–284.Young KA,Whitman JM, Turner GF, 2009. Secondary contact dur<strong>in</strong>g adaptive radiation: A community matrix for Lake Malawi cichlids. J.Evol. Biol. 22: 882–889.Zimmerer EJ, Kallman KD, 1989.Genetic basis for alternative reproductive tactics <strong>in</strong> the pygmy swordtail Xiphophorus nigrensis. Evolution,43: 1298–1307.18
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>Fig. 1 Disruptive selection <strong>in</strong> a cont<strong>in</strong>uous trait (Left) and relationship between <strong>male</strong>-<strong>male</strong> aggression and the frequency <strong>of</strong>pre-exist<strong>in</strong>g color-morphs (Right)Due to high cost <strong>of</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> among the most common phenotypes (purple), extreme phenotypes (blue and red) are favored, chang<strong>in</strong>gthe population from a unimodal to a bimodal trait distribution. <strong>The</strong>se color-morphs could have evolved either through disruptive selection <strong>of</strong> acont<strong>in</strong>uous trait (see left) or through a discrete phenotypic shift caused by e.g. a s<strong>in</strong>gle mutation caus<strong>in</strong>g a change <strong>in</strong> coloration. <strong>The</strong> cost <strong>of</strong><strong>in</strong>trasexual antagonism is expected to be <strong>in</strong>verse to the frequency <strong>of</strong> a given color-morph if there is greater aggression among <strong>male</strong>s <strong>of</strong> similarphenotypes. Such negative frequency-dependent selection predicts stable coexistence <strong>of</strong> two or more color morphs.19
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>Fig. 2 Study systems where <strong>male</strong> aggression has been suggested to be <strong>in</strong>volved <strong>in</strong> population divergence and <strong>speciation</strong>Top left: Pied flycatcher <strong>male</strong>s. Left, black and white <strong>male</strong>. Right, Grey/brown <strong>male</strong> plumage coloration, which is more common and pronounced <strong>in</strong>sympatry with the closely related collared flycatcher. This is likely due to less aggressive <strong>in</strong>teractions with collared flycatchers. Photo credit JavierBlasco Zumeta. Top right: Red and blue morph <strong>of</strong> Lake Victoria Pundamilia cichlid fishes. Colour differences, and differences <strong>in</strong> local species- andcolour-morph composition affect <strong>male</strong> <strong>contest</strong> <strong>competition</strong>. Photo credit Peter Dijkstra. Bottom left: Damselflies <strong>in</strong> the genus Calopteryx show<strong>in</strong>gvariation <strong>in</strong> w<strong>in</strong>g dots <strong>in</strong>volved <strong>in</strong> <strong>male</strong> expression <strong>of</strong> aggression. Photo credit Erik Svensson. Bottom right: Strawberry poison frogs from twopopulations with extremely different colour-morphs. Colouration co-varies with <strong>male</strong>-<strong>male</strong> aggression, which could be important, both as a possibledriver <strong>of</strong> phenotypic divergence, and through effects on reproductive isolation.20
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>Fig. 3 Fitness effects and trait optima for a trait <strong>in</strong>volved <strong>in</strong> <strong>male</strong> <strong>contest</strong> <strong>competition</strong> <strong>over</strong> <strong>mates</strong>Suggested fitness effects caused by natural (grey l<strong>in</strong>es) and sexual selection (black l<strong>in</strong>es) <strong>of</strong> a given trait value. <strong>The</strong> first graph shows a referencescenario <strong>of</strong> natural and sexual selection with the trait optima (0). <strong>The</strong> second and third graph each show two scenarios <strong>of</strong> changed fitness effectsthrough changed sexual and natural selection respectively. <strong>The</strong> second graph shows (A) a scenario where the level <strong>of</strong> the trait value does not alterfitness effects through sexual selection and (B) an <strong>in</strong>creased benefit <strong>of</strong> a higher trait value. <strong>The</strong>se situations could represent e.g. differences <strong>in</strong>population densities where (A) low densities lead to <strong>male</strong> <strong>competition</strong> not affect<strong>in</strong>g the reproductive success and (B) the opposite. <strong>The</strong> third graphshows scenarios <strong>of</strong> <strong>in</strong>creased (C) and decreased (D) cost from a higher trait value, through natural selection. This situation could represent e.g.<strong>in</strong>creased (C) or decreased (D) predation pressure on conspicuous behaviors or ornaments. Trait optima for each scenario are shown by arrows andcorrespond<strong>in</strong>g letters <strong>in</strong> the grey box.21
QVARNSTRÖM A et al.: Male aggression and <strong>speciation</strong>Fig. 4 Feedback loop between reproductive character displacement and ecological character displacement at secondarycontact between two young species exhibit<strong>in</strong>g <strong>male</strong> resource defence breed<strong>in</strong>g systemsWhen allopatric populations experience different environments (a) some degree <strong>of</strong> asymmetry <strong>in</strong> <strong>male</strong> resource defence abilities is likely to evolve.At secondary contact (sympatry) such asymmetries <strong>in</strong> <strong>male</strong> resource defence abilities may facilitate reproductive character displacement and habitatsegregation (b), which <strong>in</strong> turn cause an <strong>in</strong>creased asymmetry <strong>in</strong> <strong>male</strong> resource defence ability (c).22