temporaldynamics and communication of...

TEMPORAL DYNAMICS AND COMMUNICATION OFWINNER-EFFECTS IN THE CRAYFISH

ORCONECTES RUSTICUS

by

DANIEL A BERGMAN12) CORINNE P KOZLOWSKI123)JEREMY C MCINTYRE12) ROBERT HUBER2) ALISDAIR G DAWS2)

and PAUL A MOORE1245)

(1 Laboratory for Sensory Ecology Department of Biological Sciences Bowling Green StateUniversity Bowling Green Ohio 43403 USA 2 J P Scott Center for Neuroscience Mindand Behavior Department of Biological Sciences Bowling Green State University BowlingGreen Ohio 43403 USA 3 Department of Neurobiology and Behavior Cornell University

Ithaca New York 14853 USA)

(Acc 23-V-2003)

Summary

A variety of factors in uences the formation of hierarchical structures and can include analtered aggressive state an ability to physically dominate and previous agonistic experienceUsing male Orconectes rusticus we tested the duration of the winner effect by varying thetime between a winning encounter and a subsequent encounter by a 20 40 or 60-minuteinterval Varying the time between the two ghts signi cantly altered the probabilities ofinitiating ght behaviour and of winning a ght A cray sh with a 20-minute delay betweenits winning experience and its subsequent ght was signi cantly less likely to initiate ghtbehaviour and signi cantly more likely to win its next ght than was an animal whose next ght was delayed for 40 or 60 minutes We then investigated whether the dynamics of thiswinner effect were in uenced by perception of odour signals during agonistic interactionsby blocking the chemo- and mechanoreceptors on the antennae and antennules to preventreception of relevant cues communicating social status Individuals ghting an opponentwith this loss of sensory information were signi cantly more likely to initiate a ght butthen escalated at a slower rate to a higher ght intensity level In addition individuals had adecreased chance of winning an agonistic bout against an opponent deprived of sensory inputfrom the antennae and antennules

4) Corresponding Authorrsquos e-mail address pmoorebgnetbgsuedu5) This project was funded by a BGSU TIE grant and NSF AB-9874608 and IBN-0131320to PAM

copy Koninklijke Brill NV Leiden 2003 Behaviour 140 805-825Also available online -

806 BERGMAN KOZLOWSKI MCINTYRE HUBER DAWS amp MOORE

Introduction

Aggression plays an important role in the lives of many animals (Wilson1975) Activities relating to aggressive acts speci cally ght behaviour aretermed agonistic interactions (Scott 1951) It is through these types of in-teractions that dominance hierarchies are established When a high socialstatus is obtained through these interactions the status confers an individualwith many advantages including increased access to food mates and shel-ter while decreasing access for subordinates The use of dominant status toachieve a higher tness is found in many invertebrate (Bell amp Gorton 1978Trunzer et al 1999) and vertebrate taxa (Tilson amp Hamilton 1984) Oneinvertebrate example the American lobster Homarus americanus acquiresshelters by defeating opponents in agonistic bouts and thereby increases itsaccess to available food and mates (Hyatt 1983 Atema 1986 Cromartyet al 1999) Because of such bene ts dominant individuals experience in-creased tness over subordinate individuals (Wilson 1975) Aggression canbe costly and possibly injurious but when used in an appropriate contextualmanner can be quite invaluable to acquiring these bene ts

Crustaceans particularly cray sh have been used as a model system tostudy the formation and characteristics of dominance hierarchies becauseof the ritualized nature of their agonistic bouts (Bruski amp Dunham 1987)the formidable weaponry involved (Garvey amp Stein 1993) and the useof sensory information during such encounters (Zulandt-Schneider et al1999 2001) An agonistic encounter typically begins when an individualapproaches a potential opponent It may then progress to a series of threatdisplays and if neither individual retreats nor escapes progress to a gradualincrease in ght intensity (Huber amp Delago 1998) In the most intense inter-actions cray sh rely on periods of unrestrained ghting where each individ-ual attempts to injure an opponent by grasping at chelae legs or antennaeFighting will continue until one individual retreats usually signi ed by atail ip away from the opponent (Bruski amp Dunham 1987) An individual isconsidered subordinate when it begins to retreat consistently thus allowingunimpeded access to resources for the victor

Asymmetries in ghting ability or resource holding power (RHP) mayexist through some intrinsic feature or extrinsic circumstance that favors onecontestant (Maynard Smith amp Parker 1976) Asymmetries serve as accuratepredictors of dominance during interactions between pairs of cray sh and

TEMPORAL DYNAMICS amp COMMUNICATION OF WINNER-EFFECTS 807

may include physical body size (Bovbjerg 1953 1970 Rubenstein amp Ha-zlett 1974 Berrill amp Arsenault 1984 Pavey amp Fielder 1996) and chelaesize (Garvey amp Stein 1993 Rutherford et al 1995) Other asymmetriessuch as prior residence (Peeke et al 1995) differing ght strategies (Guiasuamp Dunham 1997) and previous social experience (Rubenstein amp Hazlett1974) or history in agonistic encounters (Daws et al 2002) determine theoutcome of agonistic interactions

An animalrsquos previous winning or losing experiences may affect the out-come of subsequent interactions namely the probability of winning the nextagonistic interaction (Dugatkin 1997 Hsu amp Wolf 1999) Individuals thatexperience a win during an agonistic interaction are more likely to win thenext encounter against both familiar and naive opponents This increasedlikelihood of winning is termed the winner effect The converse is also truefor the loser of an encounter resulting in an increased probability of losinga subsequent ght In general both winner and loser effects have been ob-served in several species of sh including the blue gourami Trichogastertrichopterus (Frey amp Miller 1972) paradise sh Macropodus opercularis(Francis 1983) and the green swordtail sh Xiphorus helleri (Franck ampRibowski 1987 Beaugrand et al 1991) birds (Jackson 1991) and insects(Moore et al 1988 1997 Whitehouse 1997) However winner effects donot always occur for all animals (Francis 1983 Beacham amp Newman 1987Schuett 1997)

Moreover winner and loser effects are time dependent In some speciesa prior losing experience causes an individual to lose subsequent encountersafter as long as 15 to 24 hours (Francis 1983 1987 Beaugrand amp Zayan1985 Beacham amp Newman 1987 Bakker et al 1989) In experimentswith pumpkinseed sh this winning effect was present immediately after aprevious win and after a time interval of 15 minutes After 1 hour howeverprevious winners no longer held an advantage against opponents (Chaseet al 1994) However Hsu amp Wolf (1999) demonstrated that winner effectsin a cyprinodont sh can last for at least 48 hours Winner effect durationsare variable and often species speci c

While the proximate causation of winner effects is not well understoodseveral factors may be contributing to a previous winner having an increasedpropensity to win future interactions Winner effects may be a result of intrin-sic changes or changes in the internal motivation of the winning individualThese changes could involve alterations in the way an individual perceives


the ghting ability of an opponent For example animals may associate thesize of an opponent with its ghting ability (Otronen 1990) or alternativelyan animal may assess its own resource holding potential (RHP) compared tothe population distribution of ghting abilities (Parker 1974 Whitehouse1997) The causation of winner effects could also include changes in sero-tonin levels that have been shown to produce both heightened aggressivestates and an increase in agonistic behaviours (Antonsen amp Paul 1997) Ifintrinsic causation is responsible for the results of agonistic interactions ananimal that has had a prior winning experience may function as a lsquosuccess-fulrsquo ghter and thereby ght more readily in future agonistic interactionsThese effects are likely caused by changes in internal motivation but mayalso be mediated by a communication of status to which a potential oppo-nent interprets and responds

Alternatively the winner effect may result from extrinsic changes or exter-nal characteristics Such changes could include recognition of a conspeci cwith a heightened aggressive state (Copp 1986) In Crustacea recognition ofaggressiveness could be accomplished visually by examining posture (Thorpamp Ammerman 1978 Winston amp Jacobson 1978 Bruski amp Dunham 1987)or by detecting a change in the physiological state of the opponent that is ex-pressed externally by chemical cues For instance serotonin plays differentroles in the neurochemistry of dominant and subordinate individuals (Yehet al 1996 1997) This intrinsic alteration could possibly be expressed ex-trinsically through chemical signals such as metabolites that are released byan individual with a previous winning experience during an encounter in theenvironment (Zulandt-Schneider et al 1999 2001)

Detection of chemical signals through the use of antennae and antennuleshas been well documented in crustaceans In lobsters chemical cues releasedwith the urine are important in status andor individual recognition (Kara-vanich amp Atema 1998) and mating behaviour (Snyder et al 1992 1993)Karavanich amp Atema (1998) also demonstrated that chemical cues in uencethe progression and outcome of agonistic encounters in lobsters When in-dividuals were deprived of the ability to detect odours by obstructing thechemoreceptors (anosmic) they were unable to recognize familiar oppo-nents This lack of individual recognition altered both the intensity and dura-tion of the repeated agonistic encounters producing second ghts that weresimilar to the rst interaction between the individuals Zulandt-Schneider


et al (2001) demonstrated that cray sh determine the social status of op-ponents by using chemical cues released with the urine of the cray sh Or-conectes rusticus When these cues were removed agonistic battles werelonger and reached higher intensity levels (Zulandt-Schneider et al 2001)These studies indicate that information pertaining to the behavioural state ofan individual can be transmitted through chemicals released with the urineand can in uence the behaviour of its opponent Zulandt-Schneider et al(1999) have also shown that Louisiana red swamp cray sh Procambarusclarkii predominantly rely on urine chemical cues to communicate domi-nance status

The goals of this study are two-fold In male cray sh (Orconectes rusti-cus) we rst explore the temporal dynamics of winner effects and then ex-amine the roles of the antennae and antennules in receiving communicationduring agonistic encounters First by focusing on ght dynamics we aim todetermine whether an individualrsquos probability of initiating and winning ghtschanges as a function of the time elapsed between a winning experience anda subsequent encounter Secondly by removing the ability of an opponentto detect chemical and mechanical cues we tested whether the antennae andantennules in uence the ght dynamics of agonistic encounters

Materials and methods

Animals

All cray sh used in this study were male Orconectes rusticus collected from the PortageRiver near Bowling Green Ohio All individuals consisted of intermolt form I males withfully intact appendages (Avault amp Huner 1985) Animals were tactilely and socially but notchemically isolated in individuallyventilatedplastic containers (1778 mm ID) maintained ina tank with re-circulatingwater at a constant temperature (23plusmnC) and a 14h 10h (light dark)cycle for a minimum of one week prior to the experiments Cray sh were fed 1 rabbitfood pellet three times per week Descriptive statistics (mean sect SEM) for a total of 234cray sh used in this study are carapace length (311 sect 54 mm) chelae length (297 sect77 mm) and weight (113sect58 g) Cray sh were marked individually on the dorsal carapaceusing whiteout (Liquid Paperreg) and were used only once in the course of the study At theconclusion of this study the animals were returned to the branch of the Portage River wherethey were collected

Time dependency of the winner-effect

Fight trial

In order to determine the time dependence of a winner effect three cray sh were used foreach trial Each cray sh was placed in a separate compartment of a ght arena (Fig 1) that


was constructed of opaque Plexiglas (40 pound 40 pound 14 cm) The arena was divided into fourquadrants of equal size (20 pound 20 pound 14 cm) using opaque retractable walls The arena was lled with 15 liters of de-chlorinatedwater After a fteen-minute acclimation period a wall(1 in Fig 1) was removed and cray sh (A) and (B) were allowed to interact Cray sh (B) wasmeasured to be a minimum of 25 smaller in carapace length (241sect35 mm) chelae length(202 sect 46 mm) and body mass (483 sect 22 g) than cray sh (A) (carapace 330 sect 41 mmchelae 325 sect 62 mm weight 1319 sect 52 g) All three measurements were taken to ensurethat cray sh (A) would have a winning experience and would thus emerge as the dominantindividual at the end of the ght These two individuals were allowed to interact for a singleencounter ie until one individual retreated or tail ipped away from its opponent creatingdominant-subordinate relationship Simultaneously wall (2 in Fig 1) was removed and thefuture opponent cray sh (C) (carapace 330 sect 41 mm chelae 325 sect 60 mm weight1315 sect 53 g) was allowed to explore the same amount of area as cray sh (A) and (B) Thiswas done to ensure that all animals had the experience of seeing the walls retracted Afterthe rst encounter concluded all three animals were returned to their respective startingpositions Forcing the animals into their original compartments was done so that all threecray sh received the same handling treatment

After waiting periods of either 20 40 or 60 minutes a wall (3 in Fig 1) in the test arenawas removed and cray sh (A) and (C) were allowed to interact Cray sh (A) and (C) weresize matched to within 10 of carapace length chelae length and body mass (see previousmeasurements) These two individuals were allowed to interact for a single encounter ieone retreated or tail ipped from its opponent The test arena was cleaned after each trial withdeionized water and re lled for subsequent trials Ten trials were conducted for each timeinterval using a total of 90 cray sh

Fig 1 The ght arena with retractablewalls Fight subject (A) was the test subject in whichdominance was established by defeating the smaller cray sh (B) Fight subject (C) was theindividual that (A) fought after a time interval of 20 40 or 60 minutes in the temporal effectexperiment and was also the individual with impaired chemo- and mechanoreceptors in the

blocked antennae and antennules experiment


The role of the antennae and antennules in winner effects

Chemo- and mechanoreceptor blocking technique for antennae and antennules

In the second part of this experimentwe examined the roles of chemical and mechanical com-munication in relation to the antennae and antennules in the maintenance of winner effectsFirst cray sh (A) and (C) were placed in ice for ten minutes and then removed and restrainedfor fteen minutes (see Fig 1) During the restraint period animals designated cray sh (C)received a superglue (Duroreg containingcyanoacrylate)application to the antennae and anten-nules in the experimental trials This would later serve to block this individualrsquos chemo- andmechanoreceptors and subsequently inhibit communication involving the antennae and an-tennules between cray sh (A) the cray sh with a previous winning experience and cray sh(C) The antennae and antennules of cray sh (A) were brushed with a Q-tip dipped in deion-ized water and also received a small dot of superglue on the carapace The brushing was doneto expose all individuals to a similar tactile treatment of their antennaeantennules and theglue was applied to control for the presence of the superglueodour Both glue and water wereallowed to dry before the animals were placed in their respective tanks The control groupsfor (A) and (C) did not receive any of the experimental manipulations A behavioural assaywas performed by exposing the olfactory organs of cray shes (A) and (C) to a directed pulseof food odour (homogenized sh) to examine if the glue had blocked the receptors Cray sh(C) that was glued exhibited no response to the odour whereas cray sh (A) responded bygrasping at the syringe that was used to deliver the odour pulse Both cray sh respondedwhen the odour was directed at the walking legs These results indicate that deionized waterdoes not disrupt chemoreception in cray sh (A) whereas the superglue on the olfactory ap-pendages of cray sh (C) was effective at blocking the chemo- and mechanosensory abilitiesof the antennae and antennules (Moore unpublished data)

An additional experimental series was also performed that is not included in the resultsto address any concerns that the superglue adversely affects cray sh (C)rsquos behaviour As acontrol naive cray sh (A) fought naive cray sh (C) that received the superglue treatmentto examine if the glue was responsible for the experimental results shown to occur laterThe trials indicate that the glue does not signi cantly affect the probability that cray sh (C)would initiate or win an agonistic encounter (N D 12 Acirc2 D 00 p gt 005 macr D 007) Forboth initiation and winning of ghts there was an equal probability (50) that either cray shinitiated or won an interaction At the conclusion of the superglue experiment all treatedanimals were housed in aquaria until moulting that removed the superglue from the olfactoryappendages and then these animals were returned to the wild

Fight trial set-up

Fight trials were conducted in a similar manner described for the time dependent effectswith the exception of the added superglue application and the following The cray sh wereplaced in the ght arena and allowed to acclimate for fteen minutes before any interactionstook place In ght one cray sh (A) and (B) (same as temporal effect experiment) werenow allowed to interact for 5 minutes following their rst interaction to create a reinforcedwin Another signi cant change from the previous experimental protocol included a uniformtwenty-minute period between ghts one and two Following the twenty-minute period wall(3) of the test arena was removed and cray sh (A) was allowed to interact with cray sh (C)ie the individualwhose chemo- and mechanoreceptorswere blocked with superglueTwenty


trials were conducted for both control and glue treatments using a total of 120 animals In thelatter sections of the article cray sh (A) will be referred to as the lsquoprevious winnerrsquo cray sh(B) will be the lsquoloserrsquo and cray sh (C) will be either the lsquoblockedrsquo or lsquointactrsquo experimentalanimal

Fight analysis and evaluation

A video camera (Panasonic wv-CL350) positioned one meter above the test arena recordedthe trials on a VCR (Panasonic AG-1980) and displayed them on a monitor (Sony PVM-1351 G) All taped ght trials were measured using a double blind design where neitherexperimenter nor the person analyzing the tapes had access to the particular experimentalstatus of the individuals The ghts were analyzed using an ethogram modi ed in our lab(Table 1) which was based on that of Bruski amp Dunham (1987) Temporal dynamics inbehaviour were recorded including the total time for the encounter and the time it tookto reach different intensity levels The identities of initiating and retreating animals wererecorded for each encounter A winner was determined when its opponent (ie the loser)retreated or tail ipped away Instances for initiating and winning were analyzed using amultiple comparisons for proportions contingency table (q00514 D 3633) that allows fortesting analogous to the Tukey or Student-Newman-Keulstests (Zar 1999) Signi cant resultsare represented by giving a q00514 gt 3633 from the multiple comparisons test and a p lt

005 In both experiments initiation and winning were compared against an expectation ofrandom behaviour (50 chance of winning or initiating) and against treatments An analysisof time to different intensities was performed using a one-way MANOVA with a Tukey-HSDpost-hoc analysis An additional power analysis (Power D 1 iexcl macr) was included for eachstatistical test

TABLE 1 Cray sh ethogram codes (used to score ght intensity levels)

Intensity DescriptionLevel

iexcl2 Tail ip away from opponent or fast retreatiexcl1 Retreat by slowly backing away from opponent

0 Visually ignore opponent with no response or threat display1 Approach without a threat display2 Approach with meral spread threat display usually accompanied by an antennal

whip3 Initial claw use by boxing pushing andor touching with closed claws4 Active claw use by grabbing andor holding opponent5 Unrestrained ghting by pulling at opponentrsquos claws or body parts


Results

Winning effect and time dependency

A multiple comparisons for proportions contingency table was used on theproportion of ghts initiated and won by the cray sh with a previous win-ning experience cray sh (A) for the time intervals 20 40 and 60 minutes(N D 10 for each) The proportion of ghts won by the lsquowinnerrsquo after atwenty-minute time interval were signi cantly different from random forty-minute and sixty-minute interval cray sh (q D 819 q D 443 q D 819p lt 005 1 iexcl macr D 099 respectfully) (Fig 2) These cray sh won ten outof ten subsequent interactions lsquoWinnersrsquo that experienced a forty-minutetime interval won eight out of ten interactions and were signi cantly differ-ent from random and sixty-minute interval cray sh (q D 376 q D 376p lt 005 1 iexcl macr D 068 respectfully) (Fig 2) Cray sh with a sixty-minuteinterval between encounters won ve out of ten subsequent interactions andwere not signi cantly different from random

Fig 2 Proportion of ghts initiated and won by experimental cray sh 20 40 and 60minutes after a winning experience (N D 10 for each) The different letters indicate sig-ni cant difference between groups Cray sh with a 20-minute interval between ghts weresigni cantly different from all other treatments for ght initiation (q00514 gt 3633p lt 005) Cray sh with 20 and 40-minute intervals were signi cantly different from all

other treatments in winning ghts (q00514 gt 3633 p lt 005)


The proportion of ghts initiated by lsquowinnerrsquo cray sh increased alongwith the time interval between interactions Zero out of ten cray sh witha 20-minute interval between interactions initiated in their subsequent en-counters which was signi cantly different from random forty-minute andsixty-minute interval cray sh (q D 819 q D 819 q D 819 p lt 0051 iexcl macr D 099 respectfully) (Fig 2) After both the forty and sixty-minuteintervals lsquowinnersrsquo initiated ve out of ten ghts and this result was not sig-ni cantly different from random or against one another (q D 00 p gt 0051 iexcl macr D 006) (Fig 2)

The differences in time (20 40 60 minute intervals) between a priorwinning experience and a subsequent ght did not signi cantly affect thetime it took the lsquowinnerrsquo cray sh to reach different intensity levels (refer toTable 1 for intensities measured) The average time taken to reach intensitytwo was 30 sect 023 s after a twenty-minute interval 82 sect 211 s after aforty-minute interval and 31 sect 107 s after sixty minutes The averagetime taken to reach intensity three was 67 sect 045 s after a twenty-minuteinterval 91sect169 s after a forty-minute interval and 95sect225 s after sixtyminutes The average time taken to reach intensity four was 171 sect 112 safter a twenty-minute interval 219sect469 s after a forty-minute interval and136 sect 147 s after sixty minutes A MANOVA showed that the differencesin the time taken to reach intensity levels 2 3 and 4 between the cray shafter a twenty- forty- or sixty-minute time interval (N D 10 for each)were not signi cant (p gt 005 1 iexcl macr D 099) (Fig 3) Changes in thetime interval between a previous winning experience and a second ght didnot signi cantly affect the duration of the second encounter (MANOVAp gt 005 1 iexcl macr D 099)

Blocked chemo- and mechanoreceptor on antennae and antennulestreatments

Blocking the chemo- and mechanoreceptors of the antennae and antennulesof cray sh (C) signi cantly altered the outcome of the interactions (Fig 4)Cray sh (A) that had had a previous winning experience won 16 out of 20(080) of their subsequent interactions against opponents with intact chemo-and mechanoreceptors which was signi cantly more than a random distrib-ution (q D 552 p lt 005 1 iexcl macr D 091) When lsquowinningrsquo cray sh foughtagainst opponents with blocked chemo- and mechanoreceptors they won 11


Fig 3 Mean time (sect SEM) to reach ght intensities 2 3 and 4 for experimental cray shwith 20 40 and 60-minute intervals after a previous winning experience The differences inthe time taken to reach intensity levels 2 3 and 4 after a 20 40 or 60-minute time interval(N D 10 for each) were not signi cant (p gt 005 1iexclmacr D 099) using a one-way MANOVA

Fig 4 The proportion of lsquowinnersrsquo that initiated and won ghts after a subsequent win-ning experience The different letters indicate a signi cant difference between blocked un-blocked and random lsquoWinnersrsquo that fought cray sh with blocked olfactory appendages initi-ated signi cantly more than against intact opponents or random (N D 20 q00514 gt 3633p lt 005) lsquoWinnersrsquo won signi cantly more against opponents with intact olfactory ap-pendages than against blocked opponents or random (N D 20 q00514 gt 3633 p lt 005)


out of 20 (055) of their subsequent encounters which was signi cantly lessthan against the unblocked olfactory organ cray sh (q D 465 p lt 0051 iexcl macr D 062)

Blocking the olfactory organs of cray sh (C) also signi cantly alteredthe probability that the lsquowinnerrsquo cray sh would initiate an encounter Thecray sh with a previous winning experience cray sh (A) initiated signi -cantly more encounters with cray sh (C) when this opponent had receiveda glue application to its antennae and antennules than when it had not(Fig 4) When lsquowinningrsquo cray sh interacted with an opponent with chemo-and mechanoreceptors blocked with superglue it initiated an interaction withcray sh (C) in 15 out of 20 (075) of the encounters When cray sh (A)rsquos op-ponent had not received a glue treatment lsquowinnersrsquo initiated interactions in8 out of 20 (040) of the encounters Fights against olfactory blocked op-ponents showed a signi cant difference that was greater than both randomand unblocked opponents (q D 450 q D 624 p lt 005 1 iexcl macr D 079respectfully)

The mean time (sect SEM) it took for lsquowinnersrsquo to reach levels two andthree were signi cantly different when it fought against an opponent with itschemo- and mechanoreceptors blocked compared to an opponent with intactchemoreception When lsquowinnersrsquo fought against an opponent with blocked

Fig 5 Mean time (sect SEM) to reach ght intensities 2 3 and 4 for experimental cray sh(N D 20 for each) Asterisk indicates a signi cant difference using a one-way MANOVAwith a Tukey-HSD post-hoc test Time to reach intensity levels two and three were signi -

cantly different between the trials (p lt 005 1 iexcl macr D 099)


antennae and antennules it took 1719 sect 589 s to reach ght intensity leveltwo and 1783 sect 601 s to reach intensity three (Fig 5) Whereas whenlsquowinnersrsquo fought against the unblocked group it took 876 sect 275 s to reachintensity 2 and 724 sect 119 s to reach intensity 3 which was a signi cantlyshorter time interval than against blocked opponents (MANOVA p lt 0051 iexcl macr D 099) The time it took for either treatment to reach ght intensity4 was 2954 sect 574 s for encounters against blocked opponents and 3364 sect940 s against unblocked which was not signi cantly different (p gt 0051 iexcl macr D 099) (Fig 5) There was no signi cant difference found for theoverall ght duration (MANOVA p gt 005 1 iexcl macr D 099)

Discussion

Our results illustrate that winner effects in agonistic interactions betweenmale cray sh are observable after a single interaction are time dependentand are mediated by sensory information received through the antennae andantennules As time increases between a cray shrsquos winning experience inan agonistic encounter and a subsequent one its probability of winning di-minishes decreasing substantially over sixty minutes Winner effects had theopposite effect on the probability that an animal initiated a subsequent en-counter with an opponent After twenty minutes none of these individualsinitiated interactions with their opponents After forty minutes however thiseffect was no longer observed and the animalrsquos probability of initiation wasno longer different from random

While the probability of winning and initiation by an individual were al-tered by previous winning experiences the ght dynamics of the subsequentagonistic encounters were not Both the time taken to reach different ghtintensity levels and the average duration of the second encounter were notsigni cantly affected by the winner effect when the time between the twoencounters increased

Chase et al (1994) found that in agonistic interactions between pumpkin-seed sh the sh that won the previous interaction was more likely to defeatan opponent when the time between ghts decreased Winner effects wereextinguished when the time between the rst and second contest extendedto one-hour In addition Hsu amp Wolf (1999 2001) showed that prior win-ning experiences affected both the probabilities of winning and the ghting


behaviours of contestants in subsequent interactions However these winnereffects lasted for at least 48 hours These results along with ours suggestplasticity in winner effect duration across different species

An interesting nding of this study was the establishment of a winner ef-fect after a single agonistic encounter This was accomplished by allowingour experimental cray sh to interact with a signi cantly smaller opponentfor a relatively short amount of time until a dominant relationship was estab-lished After the establishment of dominance the two cray sh were imme-diately separated Since winner effects can be created after a brief agonisticencounter this provides clues to the underlying cause of these effects

The winner effects we examined may not be caused by long-term intrinsicphysiological changes in the experimental animal A mechanism of this typewould likely require repeated exposure to winning experiences Daws et al(2002) demonstrated that cray sh which experience two winning encoun-ters a day for three consecutive days have a higher likelihood of defeatingsigni cantly larger opponents in subsequent interactions than cray sh witha similar number of experiences However since we were able to establishwinner effects with only one brief agonistic interaction this implies that theeffects seen in both studies are likely caused by short-term neurochemicalchanges that result from a single winning interaction and can be reinforcedover repeated winning experiences These changes may then bring about in-trinsic changes in the subject that can change the probability of winning orbe expressed extrinsically possibly as a chemical signal to manipulate anopponentrsquos behaviour

A likely intrinsic source of change underlying winner effects may bechanges in biogenic amine levels Biogenic amines a family of chemicalsfound to be neurologically active have been shown to in uence the be-haviour of decapod crustaceans These compounds which include serotonin(Brown amp Linnoila 1990) octopamine (Kravitz 1986 1988 Adamo et al1995) norephinephrine (Barrett et al 1990) and dopamine (Nikulina ampKapralova 1992 Shively et al 1997) have all been shown to be importantin aggressive behaviour Serotonin has been shown to affect the aggressivestate of lobsters and cray sh (Edwards amp Kravitz 1997) Serotonin has alsobeen demonstrated to decrease an animalrsquos likelihood of retreat and tail ipbehaviour (Huber et al 1997 Huber amp Delago 1998) If changes in neu-rochemistry occur as a consequence of winning such as alteration in sero-tonin levels or regulatory mechanisms (production reuptake and receptor


up-regulation) then these changes may cause the observed short-term winnereffects It is likely that the neurochemistry is altered after a successful winin an agonistic encounter and that as the time between a winning encounterand a subsequent one increases these functions return to normal This couldaccount for the short-lived nature of the winner effects that were observed

Our results indicate that the dynamics of the second agonistic encounterremained the same and help to lend support to this hypothesis If the observedwinner effects were a result of a change in the information that a winner usesto assess the ghting ability of an opponent (Otronen 1990) or a changein the manner that an animal assesses its own ghting ability (Parker 1974Whitehouse 1997) one would expect that the winner would modify its ght-ing strategy in subsequent interactions This should be re ected in temporalchanges in ght dynamics such as the length of interactions and time takento reach different intensity levels Because no signi cant changes were ob-served in ght duration it is more likely that the changes in probability ofwinning an encounter are a result of changes in intrinsic neurochemistry thatare communicated extrinsically rather than changing ght strategies

The decreased tendency of cray sh to initiate interactions shortly after awinning encounter was unexpected and contrary to the results of studies onother organisms In studies with pumpkinseed sh animals with a previouswinning experience shortly before a second contest initiated the subsequentinteraction with another sh (Chase et al 1994) This phenomenon was alsoobserved in studies of winner effects of dark-eyed juncos (Jackson 1991)In contrast the second set of experiments help to clarify this anomaly Whenchemical and mechanical communication between the two individuals me-diated primarily through the antennae and antennules was prevented throughthe use of super glue the cray sh with a previous winning experience initi-ated a signi cant percentage of interactions with its glue-blocked opponentThis result demonstrates that the decreased initiation seen earlier on the partof the previous winner is probably not a result of a decreased aggressive stateRather it is a result of the rst agonistic encounter being communicated tothis new opponent When a cray sh with a previous winning experience in-teracts with a cray sh with its olfactory appendages intact this individualmay detect a dominant status signal from the previous winner and respondin defense

Recognition of status has been demonstrated in many organisms Malecockroaches use chemical cues in status recognition of potential opponents


(Moore et al 1997) and hermit crabs have been shown to recognize in-dividual status (Winston amp Jacobson 1978) It has also been demonstratedthat chemical cues contained in the urine of cray sh play a role in agonisticinteractions and that through these cues an opponentrsquos status is determined(Zulandt-Schneider et al 2001) If an opponent of the previous winner couldnot detect chemical cues used in status recognition then the opponent maynot displayed an appropriate response that would signal that the previouswinnerrsquos status had been recognized In this way a lack of signal receptionin the blocked opponent could lead to the increased tendency to initiate in-teractions on the part of the previous winner When opponents were ableto detect chemical cues from the previous winner the opponent may act ina way that portrayed a subordinate status simply from the detection of theprevious winnerrsquos chemical status and thus the previous winner would nolonger initiate ght behaviour more than by that expected from random

An increase in initiation on the part of the previous winner against acray sh with the olfactory impairment may also have been a result of areluctance to initiate by its opponent Due to a lack of mechanical or morelikely chemical cues containing information about the previous winnerrsquosstatus the opponent may have been uncertain regarding the status of thewinner and therefore did not initiate contact with the winner A decrease ininitiation with the winner could also have been a result of the opponentrsquosinability to perceive its own status or chemosensory signals Without thisinformation the opponent would have been unable to make a comparisonof its status to that of the winner and therefore may have been less likely toinitiate contact Regardless of a number of possible explanations behind thisincrease in initiation it appears that status recognition is occurring

The difference in ght dynamics observed between the cray sh with pre-vious winning experiences and cray sh with and without the chemo- andmechanosensory impairment is most likely a result of status recognition aswell Cray sh that fought individuals without the ability to use their olfactoryappendages spent a signi cantly longer time to escalate to intensity two andthree when compared to those who fought non-impaired cray sh This phe-nomenon has been observed in lobsters as well Lesioning the chemorecep-tors of lobsters was shown to lengthen the duration and increase the inten-sity of agonistic encounters (Karavanich amp Atema 1998) Zulandt-Schneideret al (2001) found that when cray sh were prevented from receiving chemi-cal signals they also fought for a longer period of time These studies along


with our results demonstrate that information gathered from chemo- andmechanoreceptors on the antennae and antennules may help to mediate thedecisions that an individual makes during an agonistic encounter When anindividual is deprived of these sensory cues it has a reduced ability to de-termine the status of its opponent and therefore may ght more intenselyand for longer periods Sensory deprivation has been shown to increase ag-gressiveness in other organisms as well When visual cues used in individ-ual recognition were removed increased aggressiveness was demonstrated inboth cray sh (Bruski amp Dunham 1987) and crabs Potmon uviatile (Van-nini amp Gherardi 1981)

Sensory information received through the antennae and antennules ap-pear to be crucial in determining the outcome of the agonistic interac-tions Encounters in which cray sh were confronted with opponents withtheir olfactory appendages blocked lost signi cantly more than those whofought against cray sh with intact mechano- and chemoreceptors This re-sult demonstrates that winner effects observed in cray sh are likely due tochemical recognition by an opponent Consequently a lack of informationreception decreased the probability that the previous winner would win theencounter against a blocked cray sh The fact that the previous winner nolonger consistently defeats a blocked opponent suggests that chemical ormechanical signals received by the antennae andor antennules are a vitalcomponent to winner effects

Because several biogenic amines have been found to play an importantrole in cray sh aggression in the past it would be advantageous to examinethe speci c role these chemicals play in the development of winner effectsThe expression of dominance could possibly be controlled by a single aminea combination of amines or based on the relative proportions of the aminespresent in the urine Whatever the controlling aspect of urine signaling isin the winner effect we can conclude from this experiment that the winnereffect is time and signal dependent and that communication of past winningexperiences appears to be a primary controlling factor of the winner effect

References

Adamo SA Linn CE amp Hoy RR (1995) The role of neurohormonal octopamine duringlsquo ght or ightrsquo behaviour in the eld cricket Gryllus bimaculatus mdash J Exp Biol 198p 1691-1700


Antonsen BL amp Paul DH (1997) Serotonin and octopamine elicit stereotypical ago-nistic behaviors in the squat lobster Munida quadrispina (Anomura Galatheidae) mdashJ Comp Physiol 181 p 501-510

Atema J (1986) Review of sexual selection and chemical communication in the lobsterHomarus americanus mdash Can J Fish Aqua Sci 43 p 2283-2290

Avault Jr JW amp Huner JV (1985) Craw sh culture in the United States mdash In Crustaceanand mollusk aquaculture in the United States (JV Huner amp EE Brown eds) AVIPublish Westport CT p 2-61

Bakker TCM Feuth-De Bruijn E amp Sevenster P (1989) Asymmetrical effects of priorwinning and losing on dominance in sticklebacks(Gasterosteusaculeatus) mdash Ethology82 p 224-229

Barrett JA Edinger H amp Siehel A (1990) Intrahypothalamicinjectionsof norepinephrinefacilitate feline affective aggression via alpha 2-adrenoceptors mdash Brain Res 525p 285-293

Beacham JL amp Newman JA (1987) Social experience and the formation of dominancerelationships in the pumpkinseed sun sh Lepomis gibbosus mdash Anim Behav 35p 1560-1563

Beaugrand JP amp Zayan R (1985) An experimental model of aggressive dominance inXiphophorus helleri (Pices Poecillidae)mdash Behav Proc 10 p 1-52

mdash mdash Goulet C amp Payette D (1991) Outcome of dyadic con ict in male green swordtail sh Xiphophorus helleri effects of body size and prior dominance mdash Anim Behav41 p 417-424

Bell WJ amp Gorton Jr RE (1978) Informational analysis of agonistic behaviour and domi-nance hierarchy formation in a cockroach Nauphoeta cinerea mdash Behaviour 67 p 217-235

Berril M amp Arsenault M (1984) The breeding behaviour of the northern temperate or-conectid cray sh Orconectes rusticus mdash Anim Behav 32 p 333-339

Bovbjerg RV (1953) Dominance order in the cray sh Orconectes virilis (Hagan) mdash Phys-iol Zool 26 p 173-178

mdash mdash (1970) Ecological isolation and competitive exclusion in two cray sh (Orconectesvirilis and Orconectes immunis) mdash Ecology 51 p 225-236

Brown GL amp Linnoila MI (1990) CSF serotonin metabolite (5-HIAA) studies in depres-sion impulsivity and violence mdash J Clin Psych 51 p 31-41

Bruski CA amp Dunham DW (1987) The importance of vision in agonistic communicationof the cray sh Orconectes rusticus mdash Behaviour 103 p 83-107

Chase ID Bartolomeo C amp Dugatkin LA (1994) Aggressive interactions and inter-contest interval how long do winners keep winning mdash Anim Behav 48 p 393-400

Copp NH (1986) Dominance hierarchies in the cray sh Procambarus clarkii and the ques-tion of learned individual recognition mdash Crustaceana 51 p 9-23

Cromarty SI Mello J amp Kass-Simon G (1999) Time in residence affects escape behav-iour in adult male American lobsters mdash Biol Bull 196 p 105-112

Daws AG Grills JL Konzen K amp Moore PA (2002) Previous experiences alterthe outcome of aggressive interactions between males in the cray sh Procambarusclarkii mdash Mar Fresh Behav Physiol 35 p 139-148

Dugatkin LA (1997) Winner and loser effects and the structure of dominance hierar-chies mdash Behav Ecol 8 p 583-587


Edwards DH amp Kravitz EA (1997) Serotonin social status and aggressionmdash Curr OpinNeurobiol 7 p 812-819

Francis RC (1983) Experimental effects on agonistic behaviour in the paradise shMacropodus opercularis mdash Behaviour 85 p 292-373

mdash mdash (1987) The interaction of genotype and experience in the dominance success of par-adise sh (Macropodus opercularis) mdash Biol Behav 12 p 1-11

Franck D amp Ribowski A (1987) In uences of prior agonistic experiences on aggres-sion measures in the male swordtail (Xiphophorus helleri) mdash Behaviour 103 p 217-239

Frey DF amp Miller RJ (1972) The establishment of dominance relationships in the bluegourami Trichogaster trichopterus (Pallas) mdash Behaviour 42 p 8-62

Garvey J amp Stein RA (1993) Evaluating how chela size in uences the invasion potentialof an introduced cray sh (Orconectes rusticus) mdash Am Nat 129 p 172-181

Guiasu RC amp Dunham DW (1997) Initiation and outcome of agonistic contests in maleform I Cambarus robustus Girard 1852 cray sh mdash Crustaceana 70 p 480-496

Hsu YY amp Wolf LL (1999) The winner and loser effect integrating multiple experi-ences mdash Anim Behav 57 p 903-910

mdash mdash amp mdash mdash (2001) The winner and loser effect what ghting behaviours are in uencedmdash Anim Behav 61 p 777-786

Huber R amp Delago A (1998) Serotonin alters decisions to withdraw in ghting cray shAstacus astacus the motivational concept revisited mdash J Comp Physiol 182 p 573-583

mdash mdash Smith K Delgado A Isaksson K amp Kravitz EA (1997) Serotonin and aggressivemotivation in crustaceans altering the decision to retreat mdash Proc Natl Acad Sci 94p 5939-5942

Hyatt GW (1983) Qualitative and quantitative dimensions of crustacean aggression mdash InStudies in adaptation The behaviour of higher Crustacea (S Rebach amp DW Dunhameds) Wiley amp Sons New York p 113-139

Jackson WM (1991)Why do winners keep winning mdash Behav Ecol Sociobiol 28 p 271-276

KaravanichC amp Atema J (1998)Olfactory recognitionof urine signals in dominance ghtsbetween male lobster Homarus americanus mdash Behaviour 135 p 719-730

Kravitz EA (1986) Serotonin octopamine and prolactin two amines and a peptide andaspects of lobster behaviourmdash In Fast and slow chemical signaling in nervous systems(LL Iversen amp E Goodman eds) Oxford Science Publ Oxford p 244-259

mdash mdash (1988) Hormonal control of behaviour and the biasing of behavioral output in lob-sters mdash Science 241 p 1755-1781

Maynard Smith J amp Parker GA (1976)The logic of asymmetric contests mdash Anim Behav24 p 159-175

Moore AJ Ciccone WJ amp Breed MD (1988) The in uence of social experience on thebehavior of male cockroaches Nauphoeta cinerea mdash J Insect Behav 1 p 157-168

Moore PJ Reagan-Wallin NL Haynes KC amp Moore AJ (1997) Odour conveys statuson cockroaches mdash Nature 389 p 25

Nikulina EM amp Kapralova NS (1992) Role of dopamine receptors in the regulation ofaggression in mice relationship to genotype mdash Neurosci Behav Physiol 22 p 364-369


Otronen M (1990) The effect of prior experience on the outcome of ghts in the buryingbeetle Nicrophorus humator mdash Anim Behav 40 p 980-982

Parker GA (1974) Assessment strategy and the evolution of ghting behaviour mdash J theorBiol 47 p 223-243

Pavey CR amp Fielder DR (1996) The in uence of size differential on agonistic behaviourin the freshwater cray sh Cherax cuspidatus mdash J Zool 238 p 445-457

Peeke HVS Sippel J amp Figler MH (1995) Prior residence effects in shelter defensein adult signal cray sh (Pacifastacus leniusculus) results in same- and mixed-sexdyads mdash Crustaceana 68 p 873-881

Rubenstein DI amp Hazlett BA (1974) Examination of the agonistic behaviour of thecray sh Orconectes virilis by character analysis mdash Behaviour 50 p 193-216

Rutherford PL Dunham DW amp Allison V (1995) Winning agonistic encounters by malecray sh Orconectes rusticus (Girard) (Decapoda Cambaridae) chela size matters butchela symmetry does not mdash Crustaceana 68 p 526-529

Scott JP amp Fredericson E (1951) The causes of ghting behaviour in mice and rats mdashPhysiol Zool 24 p 273-309

Schuett GW (1997) Body size and agonistic experience affect dominance and matingsuccess in male copperheadsmdash Anim Behav 54 p 213-224

Shively CA Grant KA Ehrenkaufer RL Mach PRH amp Nader MA (1997) Socialstress depression and brain dopamine in female cynomolgus monkeys mdash Ann N YAcad Sci 807 p 574-577

Snyder MJ Ameyaw-Akum C amp Chang ES (1992) Mating behaviour in vision-deprived American lobsters Homarus americanus mdash Mar Behav Physiol 21 p 227-238

mdash mdash mdash mdash amp mdash mdash (1993) Sex recognition and the role of urinary cues in lobsters Homa-rus americanus mdash Mar Behav Physiol 24 p 101-116

Thorp JH amp Ammerman KS (1978) Chemical communication and agonism in the cray- sh Orconectes rusticus (Girard 1852) (Decapoda Cambaridae) mdash Crustaceana 69p 668-674

Tilson RL amp Hamilton WJ (1984) Social dominance and feeding patterns of spottedhyenas Crocuta crocuta mdash Anim Behav 32 p 715-724

Trunzer B Heinze J amp Hoelldobler B (1999) Social status and reproductive success inqueenless ant colonies mdash Behaviour 136 p 1093-1105

Vannini M amp Gherardi F (1981)Dominance and individual recognitionin Potmon uviatile(Decapoda Brachyura) Possible role of visual cues mdash Mar Behav Physiol 8 p 13-20

Yeh SR Fricke RA amp Edwards DH (1996) The effect of social experience on seroton-ergic modulation of the escape circuit of cray sh mdash Science 271 p 366-369

mdash mdash Musolf BE amp Edwards DH (1997) Neuronal adaptations to changes in the socialdominance status of cray sh mdash J Neurosci 17 p 697-708

Whitehouse ML (1997) Experience in uences male-male contests in the spider Argyrodesantipodiana (Theridiidae Araneae) mdash Anim Behav 53 p 913-923

Wilson EO (1975) Sociobiology mdash Belknap Press CambridgeWinston ML amp Jacobson S (1978) Dominance and effects of strange conspeci cs on

aggressive interactions in the hermit crab Pagurus longicarpus (Say) mdash Anim Behav26 p 184-191

Zar JH (1999) Biostatistical analysis 4th ed mdash Prentice Hall New Jersey p 563-565


Zulandt-SchneiderRA SchneiderRWS amp Moore PA (1999)Recognitionof dominancestatus by chemoreception in the red swamp cray sh Procambarus clarkii mdash J ChemEcol 25 p 781-794

mdash mdash Huber R amp Moore PA (2001) Individual and status recognition in the cray shOrconectes rusticus The effects of urine release on ght dynamics mdash Behaviour 138p 137-153


Introduction

Aggression plays an important role in the lives of many animals (Wilson1975) Activities relating to aggressive acts speci cally ght behaviour aretermed agonistic interactions (Scott 1951) It is through these types of in-teractions that dominance hierarchies are established When a high socialstatus is obtained through these interactions the status confers an individualwith many advantages including increased access to food mates and shel-ter while decreasing access for subordinates The use of dominant status toachieve a higher tness is found in many invertebrate (Bell amp Gorton 1978Trunzer et al 1999) and vertebrate taxa (Tilson amp Hamilton 1984) Oneinvertebrate example the American lobster Homarus americanus acquiresshelters by defeating opponents in agonistic bouts and thereby increases itsaccess to available food and mates (Hyatt 1983 Atema 1986 Cromartyet al 1999) Because of such bene ts dominant individuals experience in-creased tness over subordinate individuals (Wilson 1975) Aggression canbe costly and possibly injurious but when used in an appropriate contextualmanner can be quite invaluable to acquiring these bene ts

Crustaceans particularly cray sh have been used as a model system tostudy the formation and characteristics of dominance hierarchies becauseof the ritualized nature of their agonistic bouts (Bruski amp Dunham 1987)the formidable weaponry involved (Garvey amp Stein 1993) and the useof sensory information during such encounters (Zulandt-Schneider et al1999 2001) An agonistic encounter typically begins when an individualapproaches a potential opponent It may then progress to a series of threatdisplays and if neither individual retreats nor escapes progress to a gradualincrease in ght intensity (Huber amp Delago 1998) In the most intense inter-actions cray sh rely on periods of unrestrained ghting where each individ-ual attempts to injure an opponent by grasping at chelae legs or antennaeFighting will continue until one individual retreats usually signi ed by atail ip away from the opponent (Bruski amp Dunham 1987) An individual isconsidered subordinate when it begins to retreat consistently thus allowingunimpeded access to resources for the victor

Asymmetries in ghting ability or resource holding power (RHP) mayexist through some intrinsic feature or extrinsic circumstance that favors onecontestant (Maynard Smith amp Parker 1976) Asymmetries serve as accuratepredictors of dominance during interactions between pairs of cray sh and














Animals



Fight trial












Fight trial set-up













Results





















Discussion






















References

















































































Animals



Fight trial












Fight trial set-up













Results





















Discussion






















References














































































Fight trial set-up













Results





















Discussion






















References















































































Results





















Discussion






















References




















































































Discussion






















References






















































































References




































































temporaldynamics and communication of...

Documents