lable at ScienceDirect

Animal Behaviour 89 (2014) 39e43

Contents lists avai

Animal Behaviour

journal homepage: www.elsevier .com/locate/anbehav

The physiological cost of courtship: field cricket song results inanaerobic metabolism

Sophie L. Mowles*

School of Biosciences, University of Nottingham, Loughborough, U.K.

a r t i c l e i n f o

Article history:Received 18 September 2013Initial acceptance 28 October 2013Final acceptance 22 November 2013Available online 14 January 2014MS. number: 13-00780R

Keywords:courtshipcricketGryllus bimaculatusmate choicerepeated signalstamina

* Correspondence: S. L. Mowles, Room C36, TheBiosciences, The University of Nottingham, Suttonborough, Leicestershire LE12 5RD, U.K.

E-mail address: sophie.mowles@nottingham.ac.uk

0003-3472/$38.00 � 2013 The Association for the Stuhttp://dx.doi.org/10.1016/j.anbehav.2013.12.014

During courtship interactions, males typically perform displays that are assumed to demonstrate someaspect of their quality. While some displays are elaborate and spectacular in appearance, others involvecomparatively simple repetitive actions. The functions of these dynamic repeated courtship displays arehypothesized to fall into two broad categories. Either the signal is repeated as a process of validation tocounter errors in transmission, or the process of repetition itself advertises the ability to bear signallingcosts. The function of any repetitive courtship display can thus be identified by investigating the natureof production costs as well as the pattern of repetition. In the present study, I investigated the function ofsignal repetition using a commonly used organism, the Mediterranean field cricket, Gryllus bimaculatus.Male G. bimaculatus court females by producing acoustic signals using rapid, repeated movements oftheir wings. I tested for the presence of energetic costs by analysing haemolymph lactate concentrationsafter time-controlled courtship interactions. Males that had performed 5 min of courtship were found tohave higher levels of haemolymph lactate relative to controls. Furthermore, those individuals producingmore rapid song had higher levels of haemolymph lactate. The song produced by courting cricketstherefore incurs significant energetic costs, and probably serves to advertise the energetic cost capacityof the calling male. Thus, I confirm that stridulation is an energetically costly signal and demonstrate thatanaerobic metabolism appears to be a significant cost of repeated courtship displays.� 2013 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.

During courtship interactions, the courting individual usuallysignals its quality as a potential mate, in its capacity to provideresources (Howard, 1978; Thornhill, 1976), parental care (Knapp &Kovach, 1991) or simply ‘good genes’ with which the courted in-dividual shouldwant to provision its young (Hamilton & Zuk,1982).Such communication may involve elaborate multimodal displaysinvolving colour, sound and motion cues to convey informationaboutmale quality to the female; these displays occur in a variety ofanimals from salticid spiders (Girard, Kasumovic, & Elias, 2011) topeacocks (Dakin & Montgomerie, 2009). The study of such systemshas received enormous interest historically, but it has only recentlybeen hypothesized that simple repetitive displays may be of com-parable importance in mate choice (Mowles & Ord, 2012). Theemphasis here is on identifying why courtship displays are per-formed repeatedly.

The functions of repetitive displays are hypothesized to fall intotwo broad categories. Repeated displays may be advantageous toanimals by ensuring that information is accurately transferred (i.e.

Gateway Building, School ofBonington Campus, Lough-

.

dy of Animal Behaviour. Published

any error in transmission is reduced by repeating the signal; seeEnquist & Leimar, 1983; Enquist, Leimar, Ljungberg, Mallner, &Segerdahl, 1990), or the repetitive nature of the display may itselfbe informative of the signaller’s ability to withstand signallingcosts. The costs of signal production may be extrinsic (i.e. circum-stantial costs), occurring as a result of producing the display, suchas reduced time for performing other activities or the risk ofattracting predators (Tuttle & Ryan, 1981; Zuk, Rotenberry, &Tinghitella, 2006), or may be intrinsic, relating directly to the in-dividual’s ability to produce the display. The display may thus belimited by the physiological capacity of the individual to producethe repeated elements of the dynamic display (Kotiaho et al., 1998;Mappes, Alatalo, Kotiaho, & Parri, 1996). Endurance capacities, or‘stamina’, may be informative in mate choice as they communicatethe quality of the signaller by demonstrating (1) its energy reservesand thus its capacity to gather resources (which are then used inthe production of display), or (2) its ability to performwell in otherecologically relevant activities (Irschick & Garland, 2001). A highdegree of stamina may confer a greater ability to disperse, searchfor resources, search for mates, beat rivals and evade predators(Leal,1999). These are qualities that are likely to confer survivabilityto potential offspring (see Mowles & Ord, 2012), and therefore re-petitive displays relying on signaller stamina should be attended toby females.

by Elsevier Ltd. All rights reserved.

S. L. Mowles / Animal Behaviour 89 (2014) 39e4340

Courtship displays are thus produced repetitively because thesignal either needs to be repeated to reduce error in transmission,in which case the signal is produced at a constant rate (Enquistet al., 1990; Enquist & Leimar, 1983; Mowles & Ord, 2012), orbecause some aspect of repetition itself contains the informationbeing communicated, in which case signal escalation is predicted(Mowles & Ord, 2012; Payne & Pagel, 1997). In the present study Iaimed to highlight the possible function of courtship signal repe-tition in a commonly used model organism, the Mediterraneanfield cricket, Gryllus bimaculatus. In this species, males court fe-males by producing repetitive acoustic signals using movements oftheir wings, termed stridulations. This courtship song is composedof a sequence of soft, discrete ‘ticks’ created by bringing the wingstogether in individual movements (Rantala & Kortet, 2003), inter-spersed with occasional trills created by vibrating the wingsrapidly. It has been shown that females prefer males that producemore vigorous courtship songs (Rantala & Kortet, 2003; Rebar,Bailey, & Zuk, 2009; Wagner & Reiser, 2000) and that these songsare energetically costly to produce as respiration increases duringcalling (Hack,1998). What remains to be seen is the extent towhichthese energetic costs drive the signalling males to their aerobiclimits and whether the accumulation of costs is related to thevigour of the display, consistent with a signal of stamina (Mowles &Ord, 2012; Payne & Pagel, 1997). We can investigate this by lookingat whether signalling males have exceeded their aerobic capacities,indicated by the accumulation of lactic acid, the metabolic by-product of anaerobic respiration.

Anaerobic respiration is an extremely costly process owing tothe production of lactic acid (and the lactate ions readily formedfrom this), which is damaging to tissues and causes muscle fatigue.Furthermore, anaerobic respiration is less efficient than aerobicrespiration as it liberates significantly less energy as ATP from eachmolecule of glucose (Sadava, Heller, Orians, Purves, & Hillis, 2008).Animals should thus try to avoid anaerobic respiration as much aspossible as, even though it may confer an additional opportunity tocontinue displaying, the resulting fatigue will limit the scope forperforming other activities postdisplay. Lactic acid has successfullybeen used as an index of anaerobic respiration in other signallingsystems (e.g. the aggressive repeated display of shell rapping inshell-fighting hermit crabs, Pagurus bernhardus, Briffa & Elwood,2001; and the threat posturing of side blotched lizards, Uta stans-buriana, Brandt, 2003). My aim in this study was thus to determinethe function of the courtship song of G. bimaculatus, whether itconstitutes an energetically costly display and what qualities thismight communicate about the male.

METHODS

Study Organisms

Mediterranean field crickets were maintained in a laboratorycolony established at the University of Nottingham, U.K., usingcrickets sourced from the University of Derby, U.K. The cricketswere kept at 28 �C in a temperature-controlled environment undera 12:12 h light:dark regime and were fed ad libitum a diet of driedrabbit food, with supplementary fresh vegetables (see Wynn &Vahed, 2004). Crickets were kept in continuous culture, and wereseparated into single-sex tanks once it was possible to identifymales and females based on the appearance of visible ovipositors,which are identifiable on females at approximately 4 weeks post-hatching. This single-sex storage method ensured that only virgincrickets were used in the experiments. The crickets used in theexperimental trials were all aged between 4 and 12 days aftereclosion. Interactions were all staged at a constant temperature of28 �C and involved only crickets free from obvious physical damage.

Behavioural Trials

Trials were conducted using three treatment groups designed toascertain the costliness of male song and the natural level ofcourtship effort required to stimulate mating. These were (1)courtship encounters involving a tethered stimulus female result-ing in prolonged song, (2) mating trials involving free female choiceand (3) control trials that did not include a female.

Courtship encounters (N ¼ 34) were staged in clear acrylic tanks(18 � 10 cm and 13 cm high) with a sand substrate, divided in halfby an opaque partition. A male cricket was placed in one half of thetank, and allowed to acclimate overnight (for at least 16 h). Twentyminutes before the observation period, a tethered female cricketwas introduced into the other half of the arena. The female had a5 cm long cotton tether attached to her pronotum by superglue, theother end of which was attached to a cardboard anchor placedbeneath the substrate. When the dividing partition was removed,the male attempted to court the female, but the tether preventedthe female from mating with the male. Each male was allowed tocourt for 5 min before the encounter was terminated. Mating trials(N ¼ 42) were staged as above, except that a male and femalecricket were placed in each tank, one on each side of the partition,and each was allowed to acclimate overnight (for at least 16 h).After the acclimation period, the opaque divider was removed andthe crickets were allowed to interact. Each interaction was termi-nated once the female had decided to mate with the male (butbefore mating occurred). This ensured that any physiological dif-ferences between individuals were a result of courtship perfor-mance, and were not due to engaging in mating. Control trials(N ¼ 31) were also staged, in which a female was not included inthe arena. As with courtship encounters, the partition wasremoved, and the male allowed to explore the whole arena for a5 min period, but in the absence of a female with which to interact.Each interaction was recorded from above using a SonyHDRXR160EB Handycam Camcorder.

Immediately after each encounter, each male was placed in a2.0 ml plastic tube and humanely killed by rapidly freezing in liquidnitrogen. Each cricket was then stored at �80 �C until it was ana-lysed for physiological costs. The widths of male head capsuleswere subsequently obtained by first photographing eachcricket alongside a scale bar using a Fujifilm Finepix S7000 cameraon the macro setting, then measuring the head capsule width viaimage analysis using ImageJ software (http://rsbweb.nih.gov/ij).This process provided a measure of the size of each animal, whichmay influence haemolymph lactate as a relationship between sizeand lactate has been observed in other organisms (e.g. Doake,Scantlebury, & Elwood, 2010). After the courtship encounters, thetethers were gently removed from each female without damagingthem, after which each female was added to the breeding colony.

Physiological Analysis

Individual crickets were removed from the freezer and allowedto defrost gradually at room temperature. A single front leg wasremoved from each cricket at the trochanter and the cricket heldover a 0.5 ml Eppendorf tube such that the resulting bead of hae-molymph (ca. 10 ml) could be captured. The haemolymph was thenanalysed for lactate using a standard laboratory kit (Trinity BiotechPLC, Bray, Co Wicklow, Ireland).

Quantifying Signal Escalation

To quantify the vigour of courtship accurately, the digital re-cordings of each encounter were played back at half-speed andscored using JWatcher version 1.0 event recording software

0.5

0.4

0.3

0.2

0.1

–0.1

0

Log

lact

ate

(mM

ol/m

l)

No female Untethered female

Tethered female

Figure 1. The mean concentration of haemolymph lactate in male crickets exposed todifferent female stimuli. The control group contained no females and resulted in nocourtship, the untethered female group allowed courtship to proceed until mate choicetook place (approximately 3 min), while the tethered female group resulted in pro-longed courtship (5 min). Error bars show SEs.

240 260 280 300 320 340 360 380

Mean intersignal component interval (ms)

1.6

1.4

1.2

1

0.8

0.6

0.4

0.2

0

–0.2

–0.4

Log

lact

ate

(mM

ol/m

l)

Figure 2. Bivariate scatterplot with fitted regression line illustrating the negativerelationship between the concentration of haemolymph lactate and the mean inter-signal component interval.

S. L. Mowles / Animal Behaviour 89 (2014) 39e43 41

(Blumstein & Daniel, 2007) to record the occurrence of individualsignal components (i.e. each individual ‘tick’ or ‘trill’, from here ontermed ‘stridulation’). The intervals between signal componentswere calculated as the time from the onset of a stridulation to theonset of the next stridulation.

To differentiate accurately the intervals between individualsignal components from those between bouts of signalling, I used alog-transformation method (Mowles & Ord, 2012; Ord & Evans,2003; Tolkamp & Kyriazakis, 1999). Assessing the frequency dis-tribution of the intervals between individual stridulations resultedin a positively skewed distribution and revealed that the inter-component interval lasted up to 600 ms. Any intervals above thisduration were deemed to be intersignal or interbout intervals. Theintercomponent intervals for each individual were thus analysed byremoving intervals of above 600 ms from the analysis.

Signal escalation was measured by correlating the duration ofthe intercomponent intervals against the time at which each in-terval ended during the interaction. The relationship obtainedindicated whether intervals decreased (signal escalation),increased (signal de-escalation) or remained constant throughoutthe interaction. The correlation coefficient between interval andtime provided a numerical index of escalation for each trial.

Statistical Methods

Two control trials included males that sang, whereas sixcourtship trials and 10 mating trials included males that did notcourt. Thesewere removed from the data set, resulting in 29 controltrials, 28 courtship trials and 32 mating trials for analysis. All datawere analysed using StatView 5.0 (SAS, Cary, NC, U.S.A.). Data thatwere not normally distributed were log10 transformed beforeanalysis to meet the requirements of the parametric tests. If log10transformation failed to achieve normality, then nonparametricstatistics were used instead. A one-factor ANOVA was used toexamine the effects of courtship displays on the haemolymphlactate concentration of the crickets. The factor included in themodel was treatment (courtship, mating or control), the dependentvariable was haemolymph lactate concentration and male headcapsule width was included as a covariate. However, there was nosignificant interaction involving the covariate, which was thusdeleted and the model recalculated. Spearman rank correlationswere used to test for relationships between the parameters of thepattern of courtship, haemolymph lactate and cricket body size asmeasured by head capsule width.

RESULTS

Physiology and Courtship

There was a significant difference in haemolymph lactate con-centration across the treatment groups (F2,82 ¼ 4.143, P ¼ 0.019;Fig. 1). Fisher’s PLSD post hoc analysis revealed that crickets thathad been courting for 5 min had higher haemolymph lactate con-centrations than controls (P ¼ 0.005), but that there was no dif-ference between crickets that had courted for a natural durationand controls (P ¼ 0.148) or between crickets that had been courtingfor 5 min and those that had courted for a natural duration(P ¼ 0.137).

There was no association between haemolymph lactate con-centration and the total number of stridulations performed in the5 min of courtship (rS ¼ 0.041, N ¼ 27, P ¼ 0.834). However, lactatewas negatively associated with the mean intersignal componentinterval (rS ¼ �0.455, N ¼ 27, P ¼ 0.020; Fig. 2). There was no as-sociation between haemolymph lactate and the level of escalationin stridulation (rS ¼ 0.119, N ¼ 27, P ¼ 0.544).

Mating Trials

In the mating trials, the mean courtship duration as measuredfrom the commencement of stridulation until the decision of thefemale to mount the male was 182.20 s. There was no associationbetween haemolymph lactate concentration and the total numberof stridulations performed (rS ¼ 0.191, N ¼ 30, P ¼ 0.305). Lactatewas not associated with the mean intersignal component interval(rS ¼ �0.284, N ¼ 30, P ¼ 0.126) or with the level of escalation instridulation (rS ¼ 0.053, N ¼ 30, P ¼ 0.775). There was no associa-tion between courtship duration (i.e. latency to mate) and malehead capsule width (rS ¼ 0.167, N ¼ 32, P ¼ 0.353).

Rate of Courtship

The majority of males in the courtship treatment escalated theirrate of stridulation as courtship progressed, while only half of those

S. L. Mowles / Animal Behaviour 89 (2014) 39e4342

in the mating trials escalated their rate of stridulation (Table 1).Two males in the courtship treatment de-escalated their rate ofstridulation (Table 1). There was no association between male sizeand the pattern of calling during courtship (see Appendix Table A1).

DISCUSSION

In the present study, male crickets producing a repetitivecourtship display (repeated stridulations) incurred significant en-ergetic production costs. Males that had been producing prolongedcourtship song had higher levels of haemolymph lactate relative tocontrols, indicating that they had experienced anaerobic respira-tion. Furthermore, males that signalled more rapidly, leavingshorter intervals between each signal component, appeared toaccumulate greater energetic costs as a consequence. Althoughthere was no significant relationship between the number ofstridulations performed and the haemolymph concentration oflactate, there was a significant negative relationship between theintersignal component (interstridulation) interval and haemo-lymph lactate. This demonstrates that it is the rate of signal pro-duction that is important in generating energetic costs.

Dynamic repetitive courtship displays such as the repeatedstridulation performed by crickets (Hack, 1998; Ketola & Kotiaho,2010) occur throughout the animal kingdom. Sometimes thesedynamic displays involve complex multimodal signals; however,many dynamic repeated courtship displays consist of simple re-petitive actions, for example the head nodding of lizards (Kelso &Martins, 2008), the foot flagging of anurans (Hartmann, Giasson,Hartmann, & Haddad, 2005) or the claw waving of fiddler crabs(Murai & Backwell, 2006). Dynamic repeated displays may servedistinct functions, which have been discussed in great detail in thecontext of aggressive signalling (Payne & Pagel, 1997), but thesesame functions can equally be applied to courtship displays(Mowles & Ord, 2012).

The simplest reason for signal repetition is to allow informationto be sent repeatedly to reduce errors made in transmission. Thisprocess is analogous to statistical sampling and is described by amodel termed the ‘sequential assessment model’ (or SAM, Enquistet al., 1990; Enquist & Leimar, 1983) in aggressive signalling (Payne,1998; Payne & Pagel, 1997). In the context of courtship, the functionof such repetition would allow a female repeated opportunities toassess the static characteristics of a signal produced by a male, suchas the appearance of an ornament. When repeated signals areproduced for such validation purposes, they are performed at aconstant rate and production costs are not expected to be related tothe rate of signalling. The assessment of signal costs is therefore notpart of the decision-making process (Mowles & Ord, 2012; Payne &Pagel, 1997). In the present study, I found that the rate of stridu-lation performed by courting males could change within a displaysequence, either escalating or de-escalating (Table 1). This is notconsistent with a signal validation function for repetition.Furthermore, that some males were found to de-escalate their rateof stridulation suggests that they may be succumbing to energeticcosts associated with performing the repeated display. Some dis-plays may be costly by necessity and it is the ability of the signallerto produce a vigorous display despite these costs that is actually the

Table 1The number of males producing stridulations at constant, escalating and de-escalating rates in the courtship and mating treatments

Constant Escalation De-escalation

Courtship trials 7 20 2Mating trials 17 15 0

information assessed by the receiver (e.g. Zahavi, 1975). It appearsthat such is the function of signal repetition in the present study. Asthe rate of the courtship display in G. bimaculatus is related to theaccumulation of energetic costs, we can conclude that males aresignalling their energetic cost threshold as a demonstration ofquality. If these costs, or more importantly, the ability to bear thesecosts, are important in mate choice, then we would expect femalecrickets to be selecting males on their ability to produce longer andmore vigorous displays. Indeed, female crickets are known to prefermales that produce longer courtship songs featuring more rapidstridulation (e.g. Rantala & Kortet, 2003; Rebar et al., 2009; Wagner& Reiser, 2000).

Vigour and physical fitness can demonstrate multiple aspects ofa male’s quality. The ability to produce an energetically demandingdisplay would demonstrate that the signaller has accumulatedsufficient energetic reserves to ‘waste’ in signal production (Zahavi,1975). Thus, the signalling individual is likely to be a successfulforager, able both to accumulate resources effectively and to evadepredators in the process of doing so. The threat of predation poses asignificant risk to crickets, the predators and parasitoids of whichcan use the acoustic signals produced in order to locate their target(Zuk & Kolluru, 1998). Thus, the ability to signal in the face of thesecosts can also demonstrate the extrinsic cost capacity of the cricketin its ability to avoid this predation risk. The ability to produce anenergetically demanding display will also demonstrate the staminaof the individual, that is, its ability to mobilize energy reservesefficiently and perform an energetically demanding activity, oftenover protracted periods (Payne & Pagel, 1996; Payne & Pagel, 1997).High stamina in the production of a demanding signal is likely tocorrelatewith the ability to perform other energetically demandingactivities such as evading predators (Leal, 1999) and beating rivals(Mowles & Briffa, 2012). These are qualities that a critical femaleshould want her offspring to inherit (but see Alatalo, Kotiaho,Mappes, & Parri, 1998). Furthermore, a male that can maintain avigorous, costly display is likely to be healthier than one thatcannot. A male that is facing physiological challenges, because ofdisease or parasites, is not likely to be able to produce such avigorous display of stamina as a healthy male (Hamilton & Zuk,1982; Tregenza, Simmons, Wedell, & Zuk, 2006). Indeed, it hasbeen shown that female crickets can detect male health throughassessment of the courtship display and that females prefer thesongs of those males that correspond to higher immunocompe-tence (Rantala & Kortet, 2003; Tregenza et al., 2006). Thus, byattending to the rate of energetically demanding displays, femalesmay be able to assess the capacity of a male to sequester energyreserves, his overall stamina and ability to perform demandingactivities such as evading predators, as well as avoiding matingwith diseased and parasitized males.

The results of the present study demonstrate how a simpledynamic repeated courtship display results in the accumulation ofenergetic costs, which are associated with the vigour of signalproduction, consistent with a display of stamina. Such dynamicrepeated displays occur widely throughout the animal kingdomand thus energetic costs such as those revealed in this study couldpotentially play a role in diverse examples of courtship signalswhere males use dynamic repeated displays.

Acknowledgments

Thanks to Mark Briffa and Karim Vahed for valuable discussionsand comments on this manuscript. Thanks also to Julietta Marquezfor technical support. I am grateful to Bob Elwood and two anon-ymous referees for their constructive comments. S.L.M. was fundedby a University of Nottingham Anne McLaren Fellowship.

Table A1Spearman rank correlation coefficient results for the rate of stridulation performedagainst male head capsule width

rS N P

Courtship trialsTotal number of calls overall �0.289 28 0.133Mean interval 0.086 28 0.657Level of escalation �0.006 28 0.975Mating trialsTotal number of calls overall 0.226 32 0.209Mean interval 0.077 32 0.669Level of escalation �0.080 32 0.655

S. L. Mowles / Animal Behaviour 89 (2014) 39e43 43

References

Alatalo, R. V., Kotiaho, J., Mappes, J., & Parri, S. (1998). Mate choice for offspringperformance: major benefits or minor costs? Proceedings of the Royal Society B,265, 2297e2301.

Blumstein, D. T., & Daniel, J. C. (2007). Quantifying behavior the JWatcher way.Sunderland, MA: Sinauer Associates, Inc.

Brandt, Y. (2003). Lizard threat display handicaps endurance. Proceedings of theRoyal Society B, 270, 1061e1068.

Briffa, M., & Elwood, R. W. (2001). Decision rules, energy metabolism and vigour ofhermit-crab fights. Proceedings of the Royal Society B, 268, 1841e1848.

Dakin, R., & Montgomerie, R. (2009). Peacocks orient their courtship displays to-wards the sun. Behavioral Ecology and Sociobiology, 63, 825e834.

Doake, S., Scantlebury, M., & Elwood, R. W. (2010). The costs of bearing arms andarmour in the hermit crab Pagurus bernhardus. Animal Behaviour, 80, 637e642.

Enquist, M., & Leimar, O. (1983). Evolution of fighting behaviour: decision rules andassessment of relative strength. Journal of Theoretical Biology, 102, 387e410.

Enquist, M., Leimar, O., Ljungberg, T., Mallner, Y., & Segerdahl, N. (1990). A test of thesequential assessment game: fighting in the cichlid fish Nannacara anomala.Animal Behaviour, 40, 1e14.

Girard, M. B., Kasumovic, M. M., & Elias, D. O. (2011). Multi-modal courtship in thepeacock spider, Maratus volans (O. P. e Cambridge, 1874). PLoS One, 6, e25390.http://dx.doi.org/10.1371/journal.pone.0025390.

Hack, M. A. (1998). The energetics of male mating strategies in field crickets(Orthoptera: Gryllinae: Gryllidae). Journal of Insect Behavior, 11, 853e867.

Hamilton, W. D., & Zuk, M. (1982). Heritable true fitness and bright birds: a role forparasites. Science, 218, 384e387.

Hartmann, M. T., Giasson, L. O. M., Hartmann, P. A., & Haddad, C. F. B. (2005). Visualcommunication in Brazilian species of anurans from the Atlantic forest. Journalof Natural History, 39, 1675e1685.

Howard, R. D. (1978). Influence of male-defended oviposition sites on early embryomortality in bullfrogs. Ecology, 59, 789e798.

Irschick, D. J., & Garland, T. (2001). Integrating function and ecology in studies ofadaptation: investigations of locomotor capacity as a model system. AnnualReview of Ecology and Systematics, 32, 367e396.

Kelso, E. C., & Martins, E. P. (2008). Effects of two courtship display components onfemale reproductive behaviour and physiology in the sagebrush lizard. AnimalBehaviour, 75, 639e646.

Ketola, T., & Kotiaho, J. S. (2010). Inbreeding, energy use and sexual signaling.Evolutionary Ecology, 24, 761e772.

Knapp, R. A., & Kovach, J. T. (1991). Courtship as an honest indicator of maleparental quality in the bicolor damselfish, Stegastes partitus. BehavioralEcology, 2, 295e300.

Kotiaho, J. S., Alatalo, R. V., Mappes, J., Nielsen, M. G., Parri, S., & Rivero, A. (1998).Energetic costs of size and sexual signalling in a wolf spider. Proceedings of theRoyal Society B, 265, 2203e2209.

Leal, M. (1999). Honest signalling during prey-predator interactions in the lizardAnolis cristatellus. Animal Behaviour, 58, 521e526.

Mappes, J., Alatalo, R. V., Kotiaho, J., & Parri, S. (1996). Viability costs of condition-dependent sexual male display in a drumming wolf spider. Proceedings of theRoyal Society B, 263, 785e789.

Mowles, S. L., & Briffa, M. (2012). Forewarned is forearmed: early signals of RHPpredict opponent fatigue in hermit crab shell fights. Behavioral Ecology, 23,1324e1329.

Mowles, S. L., & Ord, T. J. (2012). Repetitive signals and mate choice: insights fromcontest theory. Animal Behaviour, 84, 295e304.

Murai, M., & Backwell, P. R. Y. (2006). A conspicuous courtship signal in the fiddlercrab Uca perplexa: female choice based on display structure. Behavioral Ecologyand Sociobiology, 60, 736e741.

Ord, T. J., & Evans, C. S. (2003). Display rate and opponent assessment in the Jackydragon (Amphibolurus muricatus): an experimental analysis. Behaviour, 140,1495e1508.

Payne, R. J. H. (1998). Gradually escalating fights and displays: the cumulativeassessment model. Animal Behaviour, 56, 651e662.

Payne, R. J. H., & Pagel, M. (1996). Escalation and time costs in displays of endurance.Journal of Theoretical Biology, 183, 185e193.

Payne, R. J. H., & Pagel, M. (1997). Why do animals repeat displays? AnimalBehaviour, 54, 109e119.

Rantala, M. J., & Kortet, R. (2003). Courtship song and immune function in the fieldcricket Gryllus bimaculatus. Biological Journal of the Linnean Society, 79, 503e510.

Rebar, D., Bailey, N. W., & Zuk, M. (2009). Courtship song’s role during female matechoice in the field cricket Teleogryllus oceanicus. Behavioral Ecology, 20, 1307e1314.

Sadava, D., Heller, H. C., Orians, G. H., Purves, W. K., & Hillis, D. M. (2008). Life, thescience of biology (8th ed.). Sunderland, Massachusetts: Sinauer.

Thornhill, R. (1976). Sexual selection and nuptial feeding-behavior in Bittacus api-calis (Insecta: Mecoptera). American Naturalist, 110, 529e548.

Tolkamp, B. J., & Kyriazakis, I. (1999). To split behaviour into bouts, log-transformthe intervals. Animal Behaviour, 57, 807e817.

Tregenza, T., Simmons, L. W., Wedell, N., & Zuk, M. (2006). Female preference formale courtship song and its role as a signal of immune function and condition.Animal Behaviour, 72, 809e818.

Tuttle, M. D., & Ryan, M. J. (1981). Bat predation and the evolution of frog vocali-zations in the neotropics. Science, 214, 677e678.

Wagner, W. E., & Reiser, M. G. (2000). The importance of calling song and courtshipsong in female mate choice in the variable field cricket. Animal Behaviour, 59,1219e1226.

Wynn, H., & Vahed, K. (2004). Male Gryllus bimaculatus guard females to delay themfrom mating with rival males and to obtain repeated copulations. Journal ofInsect Behavior, 17, 53e66.

Zahavi, A. (1975). Mate selection e a selection for a handicap. Journal of TheoreticalBiology, 53, 205e214.

Zuk, M., & Kolluru, G. R. (1998). Exploitation of sexual signals by predators andparasitoids. Quarterly Review of Biology, 73, 415e438.

Zuk, M., Rotenberry, J. T., & Tinghitella, R. M. (2006). Silent night: adaptive disap-pearance of a sexual signal in a parasitized population of field crickets. BiologyLetters, 2, 521e524.

APPENDIX