Behavioral Ecologydoi:10.1093/beheco/arq073Advance Access publication 4 June 2010High resource valuation fuels ‘‘desperado’’fighting tactics in female jumping spidersDamian O. Elias,aCarlos A. Botero,bMaydianne C. B. Andrade,cAndrew C. Mason,cand MichaelM. Kasumovicc,daDepartment of Environmental Science, Policy and Management, University of California, Berkeley,Berkeley, CA 94720, USA,bNational Evolutionary Synthesis Center, 2024 W. Main St. (Suite A200),Durham, NC 27705, USA,cIntegrative Behaviour and Neuroscience Group, University of TorontoScarborough, Toronto, Ontario, Canada M1C 1A4, anddEvolution & Ecology Research Centre,School of Biological, Earth and Environmental Sciences, University of New South Wales, Kensington,Sydney 2052 NSW, AustraliaOpponent asymmetries often determine the probability of winning a fight in agonistic situations. In many animal systems, theasymmetries that drive the dynamics and outcome of male—male contests are related to resource holding potential (RHP) orterritory ownership. However, recent studies have shown that this is not the case among females and suggest that resourcevaluation may be more important in that context. We studied contests between the female jumping spider, Phidippus clarus, andcompared them with male–male contests in this same species. Our observations document several key differences between thesexes: Precontact and contact phases are longer in females, ritualized displays are rare in females but common among males, andfemale fights are more likely to end in injury or death. In sharp contrast with male contests, female weight and size do notcorrelate with signaling behavior, and the outcome of fights is predicted by differences in resource valuation rather than RHP.We interpret these differences in light of the different natural history of the sexes and discuss how the economics of fighting maylead to the evolution of ritualized displays in males and a ‘‘desperado effect’’ in females. Key words: divisive asymmetry,female—female competition, game theory, jumping spider, resource holding potential, resource potential value. [Behav Ecol21:868–875 (2010)]Theoretical investigations have proposed that opponentasymmetries determine the probability of winning a fightin agonistic situations (Parker 1974; Maynard Smith andParker 1976; Parker and Rubenstein 1981; Hammersteinand Parker 1982; Enquist and Leimar 1983, 1987, 1990;Riechert 1988; Leimar et al. 1991; Payne and Pagel 1996;Payne 1998). This deterministic relationship promotes theevolution of ritualized displays that help reduce the frequencyof escalated fights with stronger (larger, older, etc.) opponents(reviewed by Bradbury and Vehrencamp 1998; Maynard Smithand Harper 2003; Searcy and Nowicki 2005). In most animalsystems, empirical work on male–male agonistic interactionshas demonstrated that ritualized displays and contest dynam-ics are largely determined by the quality and condition ofopponents (Koivula et al. 1993; Jennions and Backwell 1996;Taylor et al. 2001; Taylor and Elwood 2003; Jennings et al.2004; Hsu et al. 2006; Prenter et al. 2006; Hoefler 2007; Arnottand Elwood 2008, 2009a; Brandt and Swallow 2009; Hsu et al.2009). However, recent studies show that female–female fightsin some of these same species tend to be less ritualized, riskier,and appear to be uncorrelated with resource holding poten-tial, hereon RHP (Robinson 1985; Koivula et al. 1993; Daleand Slagsvold 1995; Draud et al. 2004; Fowler-Finn and Hebets2006; Arnott and Elwood 2009b) (but see Rillich et al. 2009).For example, Draud et al. (2004) found that in Texas cichlids,male contest structure and outcome were decided on the basisof size asymmetries, whereas in female contests, it was not. Inconvict cichlids, Arnott and Elwood (2009b) found that fe-male fights involve more costly ‘‘bite’’ and ‘‘frontal display’’behaviors than do male fights. One possible explanation forthese intraspecific differences is that agonistic behavior infemales is driven by differences in resource potential valua-tion, hereon RPV rather than RHP (Draud et al. 2004; Tib-betts 2008).Theoretical work has suggested that RPV asymmetries maybest predict contest success in situations where the costs ofnot fighting at all are even higher than those of losing againstopponents with greater fighting ability (Grafen 1987). Forexample, whenever reproductive substrates are scarce, individ-uals may only be able to reproduce if they defend it against allrivals regardless of quality or condition (‘‘desperado effect,’’Grafen 1987). This scenario is known as a ‘‘divisive asymme-try’’ (Grafen 1987) because weaker (or smaller, lighter, etc.)individuals would never reproduce if they respect RHP asym-metries. Divisive asymmetries promote the evolution of fightswith higher risk of injury/death and little or no signaling.Although this scenario has been described theoretically, directempirical support is scarce (but see Plaistow and Siva-Jothy1996).Jumping spiders have been the focus of several recent studieson male–male aggression (Jackson 1980; Wells 1988; Jacksonand McNab 1989; Faber and Baylis 1993; Clark et al. 1999;Taylor and Jackson 1999; Taylor et al. 2000, 2001; Crosset al. 2006; Jackson et al. 2006; Cross et al. 2007; Hoefler2007; Elias et al. 2008; Kasumovic et al. 2009b; Kasumovicet al. 2010). Phidippus clarus is a member of this group thatroutinely engages in highly aggressive acts toward conspe-cifics. Males of this species exchange ritualized vibratory andvisual signals prior to contact and, in most cases, engage indirect contests consisting of leg fencing and grapplingAddress correspondence to D.O. Elias. E-mail: [email protected] 19 January 2010; revised 24 March 2010; accepted 11April 2010. The Author 2010. Published by Oxford University Press on behalf ofthe International Society for Behavioral Ecology. All rights reserved.For permissions, please e-mail: [email protected] at University of California, Berkeley on February 18, 2011beheco.oxfordjournals.orgDownloaded from(Hoefler 2007, 2008; Elias et al. 2008; Kasumovic et al. 2009b).In these contests, precontact bout duration is positively corre-lated with the number of vibrations emitted by the eventualwinner, and the duration of the physical contact phase is de-termined by the weight of the loser and, to a lesser extent, theweight of the winner (Elias et al. 2008). As a general