The effect of handling damage, mobility, body size, and fluctuating asymmetry on lifetime mating success of Ischnura graellsii (Rambur) (Zygoptera: Coenagrionidae)

12 Pages·2002·1.7 MB·English

by  Rivera, A Cordero

Odonatologica31(2):117-128 June 1,2002
Theeffect of handlingdamage, mobility,
body size, andfluctuatingasymmetry onlifetime
mating success of Ischnuragraellsii (Rambur)
(Zygoptera: Coenagrionidae)
A.Cordero+Rivera,F.J.Egido+Pérez andJ.A.Andrés*
DepartamentodeEcoloxíaeBioloxíaAnimal,Universidade deVigo,EUETForestal,
CampusUniversitario,ES-36005Pontevedra,Spain
e-mail: [email protected]
ReceivedMay 16,2001/Revised andAcceptedAugust15,2001
Several spp. ofodonates have been the subjectofsexual selection studies. Innon-
-territorialspeciesmostvarianceinlifetimematingsuccess(LMS)isaccounted forby
lifespanandspeciallyby thenumberofvisits,andrandom factors(likerainyweather)
canhave strong effectonreproductivesuccess.Here wepresent the study of2natural
populationsofI. graellsiibymarking-recapturemethods. Ourresultsshow that<J mating
success isrelated tobody size,mobility and handlingdamage,but not tofluctuating
asymmetry. Larger 6 6 hadgreater successin bothpopulations, aresult in agreement
with previousfindingsonthesamesp.Nevertheless,multivariateanalysesindicatethat
body lengthwasasignificantcorrelateofLMSinjustoneofthestudiedpopulations.We
estimatedamobilityindexforc?6averagingthedistancebetweenconsecutiveresightings.
Forlong-lived<J<J, wefoundapositiverelationshipbetweenmobilityandLMS.There
wasftcleareffectofleglossduringmarkingonsurvivorship,andamarginallysignificant
negativeeffecton LMS.Finally, westudiedthe effectofwingfluctuatingasymmetry
(FA) onLMS by capturingasampleofmarked individuals atthe end offield work.
ResultssuggestthatFAisnotanimportantcorrelate ofLMSinthis sp.
INTRODUCTION
Non-randomvariation in reproductive success is a prerequisite for the action of
sexualselection (BRADBURY & ANDERSSON,1987; ANDERSSON,1994). If
successful andunsuccessful individualsdifferin phenotypic variables andthere is
heritability,thenphenotypic characterswillshowanevolutionary change duetoselective
* Presentaddress:AnimalEcology,DepartmentofEcologyandEnvironmental Science,UmeaUniversity,
SE-90187Umea, Sweden
118 A.Cordero Rivera, F.J. EgidoPerez&J.A.Andres
pressures. In manyanimalspecies malemating success is apositive functionofbody
size,andthisisspeciallytrueforterritorialspecies(CLUTTON-BROCK,1988).Female
fecundity is alsopositively correlatedwithbody size, eitherbecauselarger females
produce larger clutches or becauselarger body size increases lifespan and hence
fecundity (WICKMAN&KARLSSON,1989).Body sizeis thereforeoneofthemost
commonly measuredvariablesinsexualselectionstudies.Inodonatesarecentmeta-
analysissuggests thatbodysizeispositivelyrelatedtoreproductive success interritorial
and non-territorialspecies (SOKOLOVSKAetal.. 2000, but see THOMPSON &
FINCKE,inpress).Nevertheless, thereisnoclearbiologicalreasontoexpectapositive
effectof body size inmale mating success in non-territorialspecies. In fact, some
studieshaveshownthatsmall malesare moresuccessful (BANKS & THOMPSON,
1985;ANHOLT,1991;CORDERO etal., 1997;CARCHINIetal., 2000), whichisin
agreementwithstudiesinothersmallflyinginsects (McLACHLAN&NEEMS, 1989).
Fluctuating asymmetry(FA) isdefinedasrandomdeviationsfromperfectsymmetry
inbilateralcharacters,andis supposed toberelatedtotheindividual’sdevelopmental
stability (PALMER&STROBECK, 1986).FA studiesare common in recent years.
The firstpapersreported asignificant relationshipbetweenFAandfitness, butrecent
studiesaremorelikelytoshow non-significanteffects(SIMMONSetal., 1999).This
situation is similar in odonates, because the firstFA studies showed significant
associationsbetweenFAandmalematingsuccess inlong-term(HARVEY&WALSH.
1993)andshort-termdata(CORDOBA AGUILAR, 1995),butrecent papersfailedto
findthisrelationship (LEUNG&FORBES, 1997;FORBESetal., 1997;JENNIONS,
1998;CARCHINI etal., 2000,2001).
Inthispaperwepresenttheresultsoftwo long-termstudiesinnaturalpopulations of
Ischnuragraellsii,asmallnon-territorialdamselfly commonintheIberianpeninsula.
CORDERO(1995) studiedmalemating success intwo populations ofthisspeciesand
foundevidencefornon-randommating success,butthephenotypic variablesrelated
to fitnesswere notidentified. Morerecently CORDERO (2000) using multivariate
techniques showedthatnon-linearselectionmightbemoreimportantthandirectional
selection onshaping malephenotype. Ouraimis totesttheeffectofbody size, FA,
mobility and marking effects(leg loss during handling) on reproductive success.
Furthermore, given that this species has been intensively studied, we provide
comparative dataontherepeatability ofcorrelatesofmating success.
METHODS
TwopopulationsofIschnuragraellsiiwerestudiedbymeansofmark-recapturemethods.Allindividuals
weremarkedwith ablack numberonthewing, usingapermanentmarker (StaetdlerPancolor @303S
pen).In 1995wemarkeddamselfliesatapondintheCampusoftheUniversityofVigo(UTM:29TNG2568),
andin 1996wefollowedapopulationinhabitingasmallcoastallagoonatthebeach ofBarra(Cangas,
Pontevedra,UTM:29TNG1279),bothinNWSpain.Ateverysightingofamarkedspecimen,werecorded
time,shore section (onlyatCampus),sex,age (fromthoracic coloration(CORDERO, 1987),and body
length(tothenearest 0.1 mm).Duringhandlingformarkingsomespecimenslost 1-3legs,becausewe
Lifetime matingsuccessin Ischnura graellsii 119
tried nottodamagewingsandretained individualsby theirlegs. AtCampuspopulationwecounted the
numberoflegslost,toseeiflegloosinghadaneffectonreproductivesuccess.
Duringthe periodofstudy, the Campuspond had 220 m ofperimeter andamaximum depth of
approximately1.2m. Theshoreline hadscarceornovegetation,withtheexceptionofthreesmallareas
coveredby Eleocharis and abigpatch ofPotamogeton.The communityofodonates was composedof
Cercionlindenii (verycommon), Orthetrum cancellatum(common),Sympetrumstriolatum(common),
Anaximperator(common),Ceriagriontenellum(rare),Calopteryxvirgo(twoobservations)andIschnura
pumilio(oneobservation).Fieldworkwas doneduring29daysbetween 11 Julyand 12August,with a
total of234 hofobservation(8.1h/day,range: 1-10.3h).Noobservations weremade on15 and 25July
and4and5August,due torainy weather. Markingwasmade dailyuntil 7August,butwecontinued
recapturingafter that date.The shoreline was divided into44sections of5 meach, tohave aspatial
reference formatingactivity. We estimated male mobilitybetween resightings calculatingthe average
numberofshore sections betweenconsecutive recaptures, assumingthat malesdidnotcross thepond
(CORDERO, 1995).We marked464malesand 440females,andresighted254and284respectively.
TheBarralagoonwasabout 300m2 atthe momentofthe study.Theemergentvegetationwasscarce
duetohighevaporation,andmaximum depthwasabout0.5m. Damselflies wereconcentrated onagrass
fieldattheshore,andperchedonJuncusandsomeCyperaceae.AtthishabitatI. graellsiiwasthedominant
species,butasmallpopulationofI.pumiliowasalsopresent (CORDERO & ANDRES, 1999).Other
species were Ceriagriontenellum (common),Enallagmacyathigerum(rare),Ischnura elegans (5
individuals),Cercion lindeni(6individuals),Lestesvirens(1individual),L.viridis(1individual),Sympetrum
striolatum(common),Anaximperator(common),Aeshnamixta(common)andA.affinis(oneindividual).
We visitedthepondduring31daysfrom 14Augustto17September,withatotalof161 hofobservation
(5.2h/day,range: 1-8h).Wemarked 863males(599recaptured)and766females(526recaptured).
To test theeffect offluctuatingasymmetryonmatingsuccesswecaptured31marked malesand 22
females on 12AugustatCampuspopulation.Thisallowed us tohave data onlifetime matingsuccess
(LMS)ofthissampleofindividuals andalsotomeasuretheirwinglengthwithanimageanalysissystem.
Unfortunatelythismethodmightproducebiasedestimatesbecausecapturedanimalsarenotabletoobtain
any morematings.We measured fore-and hindwinglengthondigitizedimagesofdissected wingsusing
GlobalLab 3.0 ® software. Wing lengthwas measured fromthe first antenodal tothe distal end of
pterostigma.Each wingwasmeasured threetimestoestimatemeasurementerror. Thiswasdonewitha
one-way ANOVA on absolute rigth-leftdifferences with individuals as the factor (MERILA &
BJORKLUND,1995).FollowingLEUNG&FORBES(1997)wecalculated measurementerroras(s2wjlhjn/
3)/[(s2 /3)+s2 1-100, where s2 is the variance withinsides (duetomeasurementerror) and
' LV with,in among-* wit.h.i.n ' '
s2 isthe variancebetween sides(FAvariation).
among ' '
Tocomparebetweenage groups,wedividedanimalsinto3categories;teneral(0-1daysold),young(3-
-5days) andold(morethan 5 days).Means arepresentedwith SEandsamplesize.Statistical testswere
performedwith GenStat®,SPSS®andBIOMStat®.
RESULTS
HANDLINGANDLEG LOSS
About10%ofindividuals(9.7%of413 malesand 10.9%of339females) lost one
or morelegs during marking. Thisfact hadnoeffectontheprobability ofrecapture
(62% recaptured inbothgroups). Wetestedtheeffectoflegslostonlifespan (onlyfor
resighted individuals)withgeneralizedlinearmethods(GLM), withobservedlifespan
asthedependent variable,sex as afixedeffect,numberoflegslostasarandomeffect
anddate,ageofmarking andbodylength ascovariables.Results indicateasignificant
120 A.CorderoRivera,F.J.EgidoPerez&J.A.Andres
effectofage atmarking on lifespan (p < 0.001), no effectofdateof marking (p =
0.141), body length (p = 0.612) and sex (p =0.919) and asignificant effectofthe
numberof legs lost(p = 0.018). There was no significant interactionlegs*sex (p =
0.864).The averagelifespan was 10.4±0.32(419) days forintactanimals,9.4± 1.04
(37) foranimals thatlost one legand 5.0±0.98 (12) foranimalsthat losttwo legs.
Noneof threemalesthatlostthreelegswas resighted.
A secondGLM tested theeffectofsex (fixed effect) andthenumberoflegs lost
(random) on the numberof matings, including dateof marking, lifespan and body
length ascovariables, forresighted animalsthatlivedlongenough toachievesexual
maturation.Resultsindicatesignificant effectsoflifespan (p <0.0001)and sex (p =
0.008),a marginally significant effectof the numberoflegs lost (p =0.091) butno
effectof date(p =0.965), body length (p = 0.893) or the interactionlegs*sex (p =
0.814).
MOBILITY
We estimatedthe minimumnumberof shore sections between consecutive
observationsofmaturemales.Mobility isdifficulttoestimateformalesobservedonly
afewtimes.Thereforeweconcentrated
onmalesobserved atleast6times (as
in CORDERO, 1995).Some males
showed high site-fidelity, whereas
others were observed at different
locationsatevery sighting. Theeffect
ofmobilityonLMSwas testedwitha
GLMwithPoissonerrors andloglink
(CRAWLEY, 1993),with LMSas the
response variateandmobilityindexas
the predictor variable. A significant
positiveeffectwasfound(p=0.032,n
=20males;Fig. 1).Lifespan, number
ofvisitsandbody length didnothave
Fig. 1,Therelationshipbetweenmobilityandlifetime a significant effect. This result might
matingsuccessin20malesthatwereobservedatleast be sensitive to the particular set of
6times. animals selected. Changing the
criterionto malesthat wererecorded atleast5 times (at least4movements, n= 43
males) gavenon-significantresults.
BODYSIZE
AtCampus population, therewas nocorrelationbetweenbody length and dateof
marking (males,r- -0.04,n=387,p =0.457;females,r=0.01,n=336,p=0.822) but
LifetimematingsuccessinIschnuragraellsii 121
Table I
Bodylengthcomparisonbetweenmatedand unmatedindividuals. — [mean±SE(N);P aftert-tests]
Population Matingstatus
age unmated mated P
Campus, cî teneral 25.I9±0.16(45) 25.83±0.28(15) 0.048
young 25.41±0.12(96) 25.67±0.17(60) 0.193
mature 26.19±0.U (110) 26.55±0.16(57) 0.065
ail 25.71±0,08 (252) 26.08±0.11 (135) 0.006
Campus, 9 teneral 25.73±0.27(34) 26.62±0.20(32) 0.012
young 26.25±0.20(46) 26.09±0.19(49) 0.551
mature 27.01±0.17(64) 27.17±0.16(74) 0.508
ail 26.48±0.12(148) 26.66±0.10(188) 0.229
Barra,6 teneral 27.06±0.08(160) 27.18±0.17 (30) 0.561
young 27.16±0.09(123) 27.24±0,14(85) 0.603
mature 27.11±0.08(191) 27.39±0.08 (229) 0.016
ail 27.11±0.05 (344) 27.33±0.06 (344) 0.004
Barra, 9 teneral 28.01±0.09(162) 27.92+0.14 (57) 0.589
young 28.05±0.11 (93) 27.56+0.11 (121) 0.003
mature 27.76±0.12(127) 27.82±0.11 (167) 0.701
ail 27.94±0.06 (382) 27.75+0.07(345) 0.038
apositive correlationbetweenbody length and ageatmarking (males, r =0.31, n=
383,p<0.001;females,r =0.27,n=299,p<0.001).Matedmalesweresignificantly
largerthanunmatedmales,butinfemalesthiseffectwas onlysignificantforindividuals
markedastenerals(Tab.I).AtBarrapopulation, therewasapositive correlationbetween
bodylengthanddateofmarking (males,r=0.29,n=818,p<0.001;females,r=0.23,
n=727,p<0.001)butnocorrelationbetweenbody lengthandageatmarking (males,
r=0.07,h=818,p=0.059;females,r=-0.07,n=727,p=0.072).As intheprevious
population, matedmaleswerelarger thanunmatedmales(Tab. I).
Theeffectofbody length onLMS was testedwith aGLMwith Poissonerrors and
log link, withLMS as theresponse variateand body length,lifespan andnumberof
visitsas independent variables.Duetonon-randomresidualsinthecompletesample,
werestrictedour analyses toindividualsthatwereobservedatleasttwice.Nevertheless
wecouldnotobtainamodelwithrandomresidualsforfemalesatCampus population.
Results arepresented inTableII.BodylengthhadasignificanteffectonmaleLMS at
Barrapopulation,andinallcases,thenumberofvisitstothepondwasthebestpredictor
ofLMSinmalesandfemales.
FLUCTUATING ASYMMETRY
Measurementerror was 21% and 20% forfore- and hindwing, but by using the
mean ofthreemeasurements,thecontributionofmeasurement error was reducedto
8.3%forforewingand7.8%forhindwing.Forboth wingssigned right-leftdifferences
122 A.CorderoRivera, F.J.EgidoPerez& J.A.Andres
TableII
Results ofaGLM analysis with Poissonevrrors andloglink. LMS is theresponse variable,andbody
length,lifespanandnumberofvisitsaretheindependentvariables. — [Individualsneverresightedwere
excluded]
Population Variable Coefficient (S.E.) P
Campus,â intercept -1.42(1.90) 0.457
body length 0.0205(0.0728) 0.779
visits 0.2184(0.0770) 0.005
lifespan 0.0045(0.0181) 0.804
Campus, $ intercept 1.44(1.06) 0.178
body length -0.0670(0.0396) 0.092
visits 0.2326(0.0311) <0.001
lifespan -0.0215(0.0105) 0.043
Barra, 6 intercept -4.16(1.06) <0.001
body length 0.1335(0.0380) <0.001
visits 0.1773(0.0214) <0.001
lifespan -0.0121(0.0111) 0.275
Barra, 9 intercept 0.66 (0.99) 0.507
body length -0.0407(0.0357) 0.255
visits 0,1862(0.0163) <0.001
lifespan -0.0183(0.0094) 0.051
were normally distributedaround amean of zero and positive kurtosis (Fig. 2;
Kolmogorov-Smimov test, forewing: 0.865,p =0.703;hindwing: 0.443,p =0.706).
Thisindicatesthatthedistributionisanexample offluctuating asymmetry(FA). There
was no correlationbetweensize (estimated as themean ofbothwings)andabsolute
asymmetry (forewing: r =0.003,n =53,p = 0.985;hindwing: r = 0.23,n= 53,p =
0.093),andthereforenocorrectionforallometrywas necessary(LEUNG, 1998).There
was nocorrelationbetweenabsoluteasymmetryofforeandhindwing(males:r=0.12,
n=31,p =0.516;females:r= -0.14,n=22,p =0.531).
The correlationbetweenabsoluteFA and LMS was not significant (Spearman’s
rankcorrelation, malesforewing: r =-0.30,n=31,p=0.104,hindwing: r =-0.05,n
=31,p=0.797;femalesforewing: r =-0.12,n=22,p=0.600;hindwing: r =0.05,n
=22,p=0.820). Figure3shows therelationshipbetweenFAandLMS inmalesand
females. Noneoftheindividualsincludedinthis samplelostlegs during marking.In
males,therewas atendency forareductioninFAwithincreasing success. Giventhat
LMSisalsoaffectedby lifespan,numberofvisitstothepondandbody size, wetested
theeffectofFAwithaGLMwithPoissonerrorsandloglink,withLMSastheresponse
variableandbody length, lifespan, numberofvisits andasymmetry (adding up the
asymmmetry value for both wings) as independent variables. Resultsindicatethat
maleLMSis notpredictedby theabovevariables(p>0.10inallcases,andp=0.185
forFA). In females, LMS was predicted by the numberof visits to the pond (p =
0.017),butnotby FA(p =0.810),andlifespan (p =0.066),andbody size (p=0.073)
LifetimematingsuccessinIschnuragraellsii 123
had marginally significant effects. Amodelwitha quadratic term forbody length
yielded similarresults.
DISCUSSION
Ourresults indicate thathandling during marking can haveasignificant effecton
fitnessifseverallegs are lost, whichinour experience, is not arare event with very
Fig. 2.The distribution ofFA in foreand hindwingofIschnuragraellsii.
124 A.Cordero Rivera,F.J.EgidoPerez & J.A.Andres
smalldamselflies. Loosing oneleg is probably not a greatproblem fordamselflies,
specially ifitisamedianleg. Thisisimportantforapplied studiesthatuse legsremoved
surgically from adults toobtain DNA.FINCKE & HADRYS (2001) removed one
mediantibiafromaterritorialdamselflyandfoundnoevidenceforany effectofleg
loss onfitness.Inour case thedifferencebetween6and 5-legged animalsisalsonot
significant. Nevertheless, the lossofmorethan one leg, particularly fromthe same
Fig. 3.TherelationshipbetweenFA andLMSinIschnuragraellsii.A:males;-B:females.
Lifetime matingsuccessinIschnuragraellsii 125
side,shouldsignificantly reducefeeding andperching performance, andperhaps also
theefficiencyofmatingattempts. Ourresults areclearandtoourknowledge areunique
intheliterature.FuturestudiesshouldrecordleglossifLMSistoberelatedtophenotypic
variablesinsmallcoenagrionids.
Studiesofsexualselectionindamselflieshavefoundthatlifespanisthemostimportant
correlateof maleLMS(FINCKEetal„ 1997),butrandomfactors (SUTHERLAND,
1985;SUTHERLAND, 1987), specially climaticvariables are also very important
(THOMPSON, 1997). InIschnura graellsii, CORDERO (1995)foundevidencefor
non-randomvariationinmaleLMS,butnophenotypic variablewasidentifiedasthe
targetforselection. This study shows thatmating efficiency forlong-lived males is
relatedto mobility. Itis possible thatmalesemploytwo mate-finding strategies: sit-
-and-waitand active searching, andundersomecircumstances thesecond strategy
mightbemoreprofitable. Thistraitisexactlythekindofmate-seeking abilitypredicted
by CONRAD & PRITCHARD (1992) as the target ofsexual selectionon male
scrambling damselflies.Furthermore, ANDERSSON (1994) suggested that sexual
selectionshouldfavourmaletraitsrelatedtomate-finding andmobility inscrambling
species, whichisinagreementwiththeresultsofthisstudy. Malemating efficiencyin
/. graellsiiwas shownto generatenon-linearselection on wing length inaprevious
field study (CORDERO,2000). Nevertheless, there mightbe biases introducedby
selectingonlyanimalsthatwereresighted atleast6times.Infact,foranimalsresighted
atleast5 timesthiseffectisnotsignificant,andthereforetheseresultsaretobeinterpreted
withcaution.
Ourmainobjectivewastotesttheeffectofbodysizeonmaleandfemalereproductive
success. It has beenclaimedthat in Odonates fitness and body size are positively
correlated.SOKOLOVSKA etal.(2000)performedameta-analysisof33studiesand
claimedtohave foundevidence for apositive effectofsize on malereproductive
success, specially interritorialspecies. Innon-territorialspecies, theeffectwas weak
ornonexistent.They alsopointedoutthatmorestudiesareneededbecausethedataset
issmall.Neverthelesstheresultsofthis meta-analysis are weakbecausetheyincluded
erroneousdataintheanalysis anddidnotcontrolfornon-linearselectionorphylogeny
(THOMPSON& FINCKE, in press). Taking into account thepresent work, four
populationsofI.graellsiihavebeenstudied(seeCORDERO, 1995).Malebody size
waspositively correlatedwithLMSinthreepopulations,butinallbutone, multivariate
analyseshavemiedouttheeffectofbodysizeon matingsuccess(see alsoCORDERO,
2000).Theonly exceptionistheBarrapopulation (Tab.I),whereapositiverelationship
betweenmalebodylengthandLMSisclearlysignificantonlyformalesmarkedafter
sexual maturation.Ifbody size were important in mating success, then this effect
shouldbeclearinallage groups.CORDERO(1995)foundevidenceforgreatermobility
inlarger males,butnoeffectofmobilityonLMS. Itis possible thatlarger malesare
moresuccessfulbecausethey havegreatermobility.Ontheotherhand,wehavealso
founda negative correlationbetween femalebody sizeand mating success atBarra
population. Inacloselyrelatedspecies, Ischnura elegans, threeindependent studies
126 A.Cordero Rivera,F.J. EgidoPerez&J.A.Andres
suggestthatsmall malesare moresuccessful thanlarge males(GITTINGS, 1988,as
citedby CORBET,1999;CORDERO etal„ 1997;CARCHINI etal„ 2000).Wesee
noreason forsexualselectiontofavourlargemalesinI. graellsiibutsmallmalesinI.
elegans. Therelationship betweensizeandfitnessisnoteasilydescribedwithasingle
correlation(CORDERO,2000;THOMPSON &FINCKE, inpress), and ourresults
indicatethatthereis no consistenteffectindifferentpopulations/years. Furthermore,
severalstudieshaveshownnon-linearselection (speciallystabilizing) on malebody
size in non-territorialspecies (BANKS & THOMPSON, 1985;CORDERO. 2000;
STOKS,2000).
We foundweak evidenceforthe useofFA as anindicatorof individualquality.
Although in males there was atendency for more successful males to be more
symmetrical (Fig. 3), thisrelationship was not significant. Giventhat we captured
malesattheendofthefieldseason, ourmethodtoestimateLMS mighthaveexcluded
unsuccessful short-livedmales,andthiscouldreducethepowerofourtest.Ourresults
arequestionable becauseour methodofcapturingmarkedanimalsattheendofsampling
periodimpeded us toobtainalarge sample,andalsoimpededanimalstoobtainmore
matings. NeverthelessifwingFAwere areliableindicatorofflightperformance, we
cannot explain thedifferencebetweenmales and females.Furthermore, ifFA is a
goodindicatorofquality,thenweshouldexpectFAofforeandhindwingtobepositively
correlated,butno correlationwas foundinthisstudy. Thereforeindividualsthatare
highly asymmetric forforewing are not asymmetric for hindwing and viceversa,
suggesting thatconclusionsmight dependentonthemeasuredtrait.Acommonclaim
inthe literatureis that FA is negatively correlatedwithheterozigosity, buta recent
meta-analysis suggestsaweakrelationshipbetweenbothvariables(VOLLESTAD et
al., 1999).Furthermore.FA seems not tobeareliablemeasureof“quality” inmany
otheranimals (SIMMONSetal„ 1999;BJORKSTEN etal.,2000).Arecentpaperon
Coenagrion scitulumfoundno correlationbetweenFAandheterozygosity, bodysize
andshort-termmating success (SMS), whileheterozygosity waspositively correlated
withbody sizeandSMS(CARCHINIetal., 2001).
Inconclusion,ourresultssuggestthatlifespanandthenumberofdays spentcloseto
the water explain thegreaterpart ofvariancein maleLMS inI. graellsii, apattern
typicalofnon-territorialdamselflies(FINCKEetal., 1997).Thepossible effectofFA
needs a closerexamination, specially with experimental phenotypic manipulations
(seeANHOLT, 1991).
REFERENCES
ANDERSSON,M., 1994. Sexualselection.Princeton Univ.Press,Princeton.
ANHOLT,B.R., 1991.Measuringselectiononapopulationofdamselflieswith amanipulatedphenotype.
EEvvoolluuttiioonn45; 1091-1106.
BANKS, M.J.&D.J.THOMPSON, 1985.Lifetime matingsuccessinthedamselflyCoenagrionpuella.
Anim. Behav. 33: 1175-1183.
BJORKSTEN,T„P.DAVID,A.POMIANKOWSKI&K.FOWLER, 2000.Fluctuatingasymmetryof