Odonatologica31(1):9-22 March I.2002 Survival rates, site fidelity and homingability in territorial Calopteryx haemorrhoidalis(Vander Linden) (Zygoptera: Calopterygidae) J.J.Beukema Netherlands Institute forSea Research, POBox 59, 1790AB DenBurg,The Netherlands e-mail:[email protected] ReceivedFebruary 19,2001/RevisedandAcceptedJune6,2001 InasmallisolatedpopulationalongasmallstreaminNESpain,ahighproportionof the adultspresentwereindividuallymarked. Duringsubsequentdays,their locations wereassessed bytwice-dailysurveysalongthe entire lengthofthe stream.Meandaily survivalrates inmature<J 3 and 2 2 and pre-reproductive 36 weresimilar,94%. Onlyduringthe firstdayaftermarkingweresurvival ratessignificantlylower(viz. 77 to 84% in the various groups). High proportions(around90%) ofmature 3 3 were foundtoreturn tothesame(territory)site every morningoncetheyhadoccupiedthat site for2ormoredays.Sitefidelitywaslow inpre-reproductive33andintermediate inmature 22. In adisplacementexperiment,67 territorial 33weretransferredone byonetodistantlocations(80to240malongthestream).Halfofthemreturnedtotheir originalterritory,usuallyonthesameday. INTRODUCTION Inmanyspecies ofOdonata,including theCalopterygidae, mature malesdefenda territory against conspecific malesandguard conspecific femalesovipositing within it.Insomespecies, territorialmaleshavebeenobservedtousethesamelocationday after day as aterritory, thus showing strong and persistent site fidelity. CORBET (1999)lists28species inwhichmaleshavebeenobservedforperiods of 10or more days atthesameterritorysite. ThislistincludesC. virgowithamaximumof40 days (KLOTZLI, 1971), C. haemorrhoidaliswith 19 days (HEYMER, 1973), and C. splendens with 13 days (ZAHNER, 1960),including interruptions forshortperiods, e.g. dueto adverse weatherconditions.Theserecords clearlyshow theexistenceof long-term attachmentto specific territorysites. However, they usually lackfurther information,such as frequency ofshiftstoother sitesanddependence onage.Inthe 10 J.J. Beukema non-territorialfemalesandimmaturemales, lowersitefidelity is expected, but any attempt to quantify the expected differencewithreproductive males appears tobe lacking. Moreover,dataonsitefidelityshouldbeincorporated intoastudyofpopulation dynamics,becausethechanceofencountering anindividualonsubsequent days atthe same locationis affectedby its dailysurvival rate andestimates of durationofsite fidelity shouldbecompared tothoseoflongevity.Therefore,estimatesonsurvival are indispensable inastudy onsitefidelity. Ifmalesleavetheirterritoryintheevening tospend thenightaway fromthewater (ase.g.C. haemorrhoidalismalesdo),multi-daysitefidelity shouldinvolveanability toreturn totheright spotandrecognize it; i.e.territorialmalesshould beabletouse somememoryofthecharacteristicsoftheirterritoryandofitssurroundings toefficiently return toitonsubsequent days. Inthepresentpaper,Idescriberesultsofmarking experiments inasmallpopulation of Calopteryx haemorrhoidalis(Vander Linden) living along a small stream in northeasternSpain. Severalhundredadultmalesweremarkedindividually. By daily surveys oftheentirearea occupied by the population, the presence along thestream and theexactlocationofmarkedindividualswas assessed during subsequent days. Because high proportions ofthemarkedindividualswerefrequently resighted, these observations allowednot only an estimateoftotal local daily survival rates (the proportionofanimalsthatdidnotdienordisperse fromthestream sectionsurveyed), butalsoofchancestoresightthematthesamesite.Bydisplacing territorialmalesand frequently watching their territoryand adjacent areas, theprobability andspeed of theirreturn wereestimated. METHODS Thepopulationsstudiedlivedalongasmallstreaminthefoothillsoftheeastern PyreneesinSpain,a fewkmNorthofthe villageofMontagut,betweenthetowns ofBesalutand OlotinGironaprovince.The streamiscalledTorrente delEstanyoland itslocation isindicated onamapby WASSCHER (1996).The sectionofthestreamthatwasregularlysurveyedextendedfromacascadeclosetothesourcetooneabout 650m downstream, where the streamemptiesitself into abroader valley. As thisrocky valleylacks almostanyvegetationanddriesupinthecourseofmostsummers,itisunsuitableforCalopteryxterritories. Therefore,the studiedadultpopulationappears tobe isolatedfromsuitableneighbouringhabitats,since theseweremorethan1km awayandseparatedbydrywoodlandandrocks. In summer,the widthofthestreamlocallyvariesbetween0.3and5.0m,itsdepthbetween0.1and1.0 m,butafterheavyrainthe water level maybe up to 1m higherasjudgedfromdebriswashed upin the shrubs.Thebottomischalkywithnumerousridges,slowingdownthecurrentspeedandcreatingnumerous pools.Theborders aremostlywoodedwith talltrees andshrubsalongmost ofthe650-m stretch,but sunnyspots arenumerousonandalongthewater.However,mostofthetotal watersurfaceisinthe shade morethanhalf ofthe day,leavingonlylimitedsuitableareasfor territories. Watertemperatures in the June-Augustperiodgenerallyrangedfrom14°Catthesourceandindeeppools to17°Catshallowstretches downstream,butoccasionallyrosetoover20°Catthelowerreaches.Airtemperaturesweremorevariable androseonsunnysummerdaystovaluesbetween20and30°C. However,onabout 10%ofthedaysthe skyremainedovercast during(almost)theentiredayandfew(ifany) Calopteryxappearednearthewater onsuchdays. PopulationbiologyofCalopleryxhaemorrhoidalis 11 Thepopulationsweresurveyed during12to20 successive days(interruptedonlyby afewdayswith inclementweather)ineach of7 years, aslisted inTable I.Highestnumbers ofCalopteryxwerepresent between 11a.m.and5p.m.(asalsoobserved alongthesamestreamby WASSCHER, 1996). On days with suitableweather,allpotentialterritorysites wereinspectedatleasttwice perday.The surveytrack followed the stream,nearlyentirelyby wading.As many aspossiblemales (andalso several females) werecaughtbygentlenettingandmarkedby auniquecombinationofwhiteand/orredpaintspots onthe outersidesofthewings(orontheuppersideofthebody).Whitecorrectionfluidandrednailvarnishwere applied.Numbers ofmarked individuals rangedfrom 14to 176 inthe 7years, totaling559 (Tab.I). Markingplace,sexandaroughestimateofagewerenoted. Anindividual wasnotedasimmatureifitdid notyetshowthefullmaturewingcoloration.After aboutoneminute(toallowthepaintto dry),amarked individual wasreleasedatthesameplace,whereitusuallyflewupintothetrees.Locationsatmarkingand at all subsequentencounters (calledresightings)werenoted with aprecision ofonetoafew meters, relativetocharacteristictrees andotherlandmarks indicated onamap.Themarkswerelargeenoughto allowrecognitionfromadistanceofseveralm. Toinvestigatetheir homingability, 67 marked maleswerecaughtwithin their territory,moved in a closednettoalocation atthe streambetween80and240maway,andreleased within about 10min after catching.These maleshadbeen marked severaldaysbeforeandresightedatthesamesiteforatleast4 daystoassessthattheywereownersofafixedterritory.Duringsubsequentsurveys (thefirstoneusually onthetransferday),theirpositionalongthestreamwasrecordedasfrequentlyasfeasible(usuallytwice daily). Thex2 testwasusedthroughoutthepaper toevaluatestatisticalsignificanceofobserveddifferencesin numbers. DAILYSURVIVAL In all years, most of the marked individuals were seenagain during subsequent surveys. Theresighting percentages were generally slightly higher inmalesthan in females,butthedifferencewas not significant(0.05<p< 0.1),whereasthedifference betweenmatureandimmatureindividualswas farfromsignificant (Tab.I). Table I Numbers ofprereproductive(immature)andmaturemalesandfemalesofCalopteryxhaemorrhoidalis marked and percentages (%)seenagainon atleast onesubsequentdayih each ofseven observation periods. — [Afterafewdaysofmarkingeffort,thenumbers ofmarked individuals includednearlyall trulyterritorialmalesandmostoftheother maturemales,butonlylowproportionsofthefemalesand immature individuals] Year Period Numbermarked Percentageresighted males females males females imm, mature imm. mature imm. mature imm. mature 1991 12/VII-23/VII 0 36 0 0 81 1992 13/VII-27/VII 0 32 0 4 75 25 1993 14/VII-02/VIII 4 142 0 30 50 85 83 1994 28/VI-15/VII 1 9 0 8 100 100 100 1996 19/VI-07/VII 3 102 0 27 100 89 78 1997 06/VI-25/VI 26 83 10 28 77 83 70 79 1998 12/VI-22/VI 1 7 0 6 100 71 50 Totals 35 411 10 103 77 84 70 78 12 J.J.Beukema Thedailyrecordsofthemarkedindividualswereusedtoassess minimalnumbers andpercentagesofindividualsstill aliveon successive daysaftermarking foruse incalculationsofdaily survivalrates andmean longevity. Effectsofoverlooking by theobserverduringsurveys orabsenceofthe damselflyalongthe surveytrackdueto adverse weather conditionscould partly be nullifiedby considering individuals encounteredona certainday to havebeen aliveon all foregoing days betweenthe markingdayandthe dayofobservation. Nevertheless,itshouldbekept inmindthat thesurvivalestimatesareboundtobeunderestimates,becausetheobserverwillcertainly haveoverlookedsomeindividualsbeforethe dayoftheirdeath. Thepercentagesofmarkedindividualsstillalivegradually declinedataregularrate during thefirstten days aftermarking when numberswere stillhighand estimates wererelativelyprecise (Fig. la).Thebest fittinglinesthrough thepointsshowninthis Figure mn slightly curvilinear, becoming Unearon alogarithmic plot starting from day-1.Thispointstoaconstantdailyrateofsurvivalstarting from1 dayaftermarking. Theselinesrunalmostparallel inmalesandfemales,corresponding toadailysurvival rate of94%betweendays-1 and-10 inboth sexes (Fig. la). Neitherdidthesurvival ratesofimmatureandmaturemalesdiffer(Fig. 1b).Lownumbersofmarkedimmature femalesdidnotpermit aseparatesurvivalestimateforthis group. The aboveestimateof daily survival of94% did not apply tothe first day after marking (day-0 to day-1). On this initialday, survival was substantially lowerthan 94% in all groups,viz. 84% in mature males, 78% inmature females and 77% in immaturemales.Differences betweensurvivalrates forthefirstandthesecond day Fig. 1.Calopteryxhaemorrhoidalis, minimalpercentages ofgroupsofmarkeddamselflies stillaliveon successive daysafter the markingat day-0:(a) summedobservations (addeddataofall 7 periodsof observation)on446 marked males (•) and 113females (o); — (b) observations made in 1997 on83 individuals markedasmature(•) and 26asimmature(o)males. PopulationbiologyofCalopteryxhaemorrhoidalis 13 aftermarking werestatistically significantbothin males(p<0.01, n=446 and373) and infemales (p <0.01, n= 113 and 87). Apparently, therewere extra losses of between ~ 10% and 17% immediately aftermarking. In other words, handling for markingroughly tripledthenormaldailydisappearancerateof6%.Thepresentmethods donotallowmaking adistinctionbetweenmortality,emigration,andadditionalmortality brought aboutby beinghandled. SITEFIDELITY During theevening, nightandearly morning, C. haemorrhoidaliswerenotpresent alongthesurveytrack. They wereprobablyroosting at communalroosts highabove thewater surface,invisibleinthecrownsofthetrees.They were seento descendinthe course ofthe morning,butnumberand locationof theroosts could not beassessed. Reproductive malestook a territory generally between 10 a.m. and noon. Marked individualswere usually resighted atthe samesite (often even on the sameperch) wherethey hadbeenseenontheprevious day.Alongtheregularly surveyed stretchof thestream,about60sites weresuitableforaterritory(i.e. theseplaces werefoundto be frequently occupied by a territorialmale). I definedand TableII namedsuch suitablesites witha Site fidelity(%)inmaturefemales andinvarious adult life diameter of generally several stages of males of Calopteryx haemorrhoidalis. — meters. These sites were larger [N = number of observations. Statistically significant than the small areas (afew m2 differencesbetweentwo figuresare indicatedby symbols , comparealso the even smaller placedinbetween the comparedfiguresonly asfarasthe 0.5to 1.5m2given byCORDO- fcicoamnpt,ar*is=onpis<c0o.n0s5i,de*r*ed=topb<e0m.0e1a,nainndgfu**l:*0==pno<n-0s.i0g0ni1-] BA-AGUILAR,2000)actually defendedatanytime.Thiswas a Sex and Dataof: consequence of the males’ condition 1997only allother years tendency to change his perch % (N) % (N) withinthesesitesaccording tothe Females: sunlit parts, which could shift all mature 53 (112) 60 (182) severalminthecourseoftheday. ** *** Therefore, it was inevitableto Males: recordanindividualas being at all mature 69 (296)-| 78(1682) thesame site if it was present **￿ within about 5 m from the imm.coloured 40 (40) location of the preceding O J observation. rec.mature 47 (32) *** To quantify site fidelity, I territorial >4d 92 (127) 95 (339) expressed it as a percentageof * *** allresights ofan individualthat were atthe same site (within5 last daysterrit. 77 (13) 60 (15) 14 J.J.Beukema m) asthelast observationofthat individual.The lastobservationwas usually onthe previous day,butmayhavebeenmorethanone dayearlier,e.g.ifweatherwas badon theintervening day. TableIIshows asummary oftheobservedresight percentages, separatelyfor 1997-only andforallotheryearstogether (1997was theonlyyearwith a high numberof immatureindividuals atmarking, allowing separateestimates for groupsofmalesof differentknownageandmaturitystate). As expected, mature females showed significantly lowersite fidelity thanmature males(Tab.II: toprows). Amongthe males,sitefidelity was dependent ontheirlife stage. Those withclearly immature colours andindividualsthat hadonly recently (< 3 days before) attained mature coloration(“rec. mature” in Tab. II) showed significantly lower site fidelity than fully mature males. Site fidelitywasparticularlyhigh (well above90%) inmature malesthat occupied a certain territory al- ready for several (> 4) days. However, it declined during the lasttwodaysoftheirlife, viz.toa significantly lowerpercentageof 60-77% in males that had occupied aterritoryfor >4 days during the period immediately preceding the last two days they wereresighted (Tab.II).Itwillbe meaningful that the few obser- vationsofaterritorialmalelosing aterritorialfight,werejustbefore theendoftheresighting period in thatmale. Site fidelity percentages were positivelyrelatedtothenumberof days an individual had already Fig. 2. Calopteryxhaemorrhoidalis,percentages ofindi- beenpresent ata certainsite.IfI vidualsthatwereresightedatthesamesite(i.e.within- 5m had seenareproductive maleata oftheformersite)afterhavingbeenseenthereon 1,2,3...8 ormoreprecedingdays(withoutinterruptionsofsightingsat certain siteon two or moredays othersites).Dataonthreegroupspresentedseparately:mature (without intervening days at 6 (°= 1997, 0=1996,*= 1993,□=summeddataof1991, another site), thechance thatthe 1992, 1994, and 1998;totalsofall7 years indicated by an next resighting wouldbe atthe uninterruptedline); —mature9(+anddashedline;totalsof samesite(andnotelsewhere)was all years); and - immature individuals (xand dottedline; totalsofall years). Apercentage is shownonly if >10 ca90%andevenhigher (ca95%) observationswereavailable (seenumberofobservationsgiven iftheindividualhadkeptthesame atthetopoftheFigure). territory for more days (Fig. 2: PopulationbiologyofCalopteryxhaemorrhoidalis 15 opencircles). Thesehigh site fidelitypercentagesaftertwoormore days atthesame siteweresignificantly(p <0.001) higher thanthepercentagesofca60%foundaftera preceding stay atthatsiteofonly one day.Notethatsuch one-day-only observations constituteda heterogeneous group,since thisgroupincludednotonly individualsthat hadreally stayed there foronly one daybut alsoindividualsthat couldhave been presenttherealsoonearlierdaysbutwerenotyetmarkedorobserved.Thismeansthat sitefidelity must havebeen lowerthan ca60%in mature malesthathadbeen ata certainlocationforonlyexactly one day.Sitefidelity inrecently maturedmales(those notyettruly territorial)amountedto47% (Tab.II). Asa consequenceoftherelatively low site fidelity in young(though looking fully mature-coloured) males, malesite fidelity tended toincrease inthe course ofthe season (June-August), reflecting the increaseoftheproportion offullyreproductive males. In immaturemales, thepercentagesoccurring atthesamelocationas onprevious days were not only lowerthan inmature males. Another differencewas thatthese percentagesdidnot showanincreasewithnumbersofdaysthey hadbeenseen atthis site(Fig.2:thexpoints).Though immatureindividualsdidnotshowaclearattachment toasite, almosthalfoftheresightings onlaterdays wereatroughly thesamelocation, i.e.withinabout5 m,butnotatspecificperchesandgenerallyseveralmawayfromthe watersurface(as opposed tofullyreproductive males,whowereoftenobservedatthe sameperchclosetothewatersurfacedayafterday).Generally,thereproductive males wereresighted every day (weatherpermitting), whereasoften several days elapsed betweensuccessiveresightingsofimmatureindividuals. Inmaturefemales,sitefidelitypercentageswere mostlylow(ca50%),butincreased tolevelssimilartothoseofreproductive malesonceanindividualfemalehadbeenat thesamesiteforthreeormoredays(seeFig. 2,+pointsanddashedline; thedifferences weresignificantbetween 1 andeither2,3,or4days,andbetween2and4 days).The differencesinsitefidelitybetweenmaturemalesandfemales(uninterruptedanddashed lineinFig. 2,respectively) weresignificant(p <0.01)after1 and2 days,butnotafter 3or4days. Thehighsitefidelityvaluesinmature femalesafterstays of3ormoredays atthesamesite wereduetoalownumberofindividualsthatreturnedtothesamesite dayafter day. Mostfemaleswere resighted less frequently and mostly atsites that differedfromdayto day(observations ofthesefemalesare includedintherelatively large 1-daygroupinFig. 2, cf.numbersofobservations). A high site fidelity of about90%, as observed in territorialmales, means that individualsusuallysticktothesamesiteforseveral days insuccession. Long seriesof 10ormore dayswerenot atallrare (though such serieswerefrequentlyinterruptedby oneor two days ifweatherwas bad). Such long series were observed in about50 males. Thelongest periodaterritorialmalewas seen atthesamelocationwas 20 days,but thiswillbeanunderestimate, since thelongest survey period lastedonly 20 days (in 1997, seeTab. I). 16 J.J. Beukema TableIII Numbers ofterritorial males ofCalopteryxhaemorrhoidalis transferred toanothersite atthe stream, and eitherneverseenalive onsubsequent days,orresighted attheiroriginalterritory site,at therelease site,orelsewhere (either betweentherelease and the territory site orbeyondthis stretchofthe stream) Year Transferred Never Location offirstresightaftertransfer number resighted Territory Release Between Beyond site area 1992 11 2 6 3 0 0 1993 35 6 15 6 5 3 1996 21 2 9 4 4 2 Totals 67 10 30 13 9 5 HOMING ABILITY By farmost(57) ofthe67transferred maleswereresighted atsubsequent surveys along thestream. Theloss of10 individualsappearstobesomewhatinexcess ofthe normaldailymortalityof6%, buttheextrahandling (catchingandmoving) mayhave enhanced the disappearance rate as observed immediately afterinitialmarking (see above). Abouthalf(30) ofthe57survivors amongthetransferredmaleswereencountered afterwards atthe site of theiroriginal territory (Tab. IB). Quick returns, already completed on the dayofthetransfer, were observedin 19males.Theproportion of quickreturns wascertainlyunderestimated,since someofthemoveswouldhavetaken place afterthelastsurvey ofthe day. Nearly halfofthe27individualsthatdidnot returntotheirterritorywereobserved tostay atthereleasesite.Many oftheotherswereresighted somewherebetweenthe releaseandtheoriginalterritorysite,but5 maleshadmovedintothe“wrong” direction (Tab. III). Many moremaleshad movedintothe“right”thanthe “wrong” direction, viz. 30+9= 39versus 5. Thus,a hypothesis thattheirreturn flightresultedinanet movementintoarandomdirectioncanberejected (p<0.001,signtestwithn=44). Resultsofupstreamtransfersdidnotdifferfromdownstreammoves: returns tothe original territory were 16outof36 upstreammoves and 14out of 31 downstream moves. Thedistanceofthemovesmayhaveaffectedthe return success. Aftershortmoves (80-140 m),24 outofthe49transfersresultedinasuccessfulreturn,whereasthiswas soinonly6outofthe 18longer moves(160-240 m).Thedifference,however,wasfar fromsignificant (p >0.1). It shouldbe noted(1) thateventheshortest distancesof about80m did neverallowa directviewtothe territorysite (direct viewsalong the stream were usuallylimitedto< 10m as aconsequenceofdensevegetation ofshrubs innearlytheentirevalley)and(2)thatseveralotherterritorysites werelocatedbetween thesitesofthe releaseandtheoriginal territory. PopulationbiologyofCalopteryxhaemorrhoidalis 17 DISCUSSION SURVIVAL Calopterygid damselfliesarewellsuitedforpopulation studiesbyindividualmarking since they are easily caught andresighted. Thepresent population ofCalopteryx haemorrhoidaliswas sufficiently small(numbering atmostafew hundred)andlived inasufficientlyconfinedareatobeefficientlysurveyed, allowing high proportionsof allindividualstobemarkedandhigh chancestoresightmarkedindividuals.Because of its isolated habitat,emigration rates will havebeenlow.Theobserved high value (94%) ofmean daily survivalrateexcludes anyhigh rate ofemigration. Themean survivalrateobservedinthepresent study (0.94perday) was similarto rates observedinothercalopterygid species, ranging from0.70to0.98per day(Tab. IV).Extremelylowandhighvaluesobservedbysomeauthorsdeservecloserinspection. Thelow(0.70and0.77)figuresinC. aequabilis, reportedby CONRAD&HERMAN (1990) werefoundin apopulation with a high proportion ofmigrants andtherefore mayreflecta high daily rate ofemigration rather thanmortality. Theexceptionally highfigureof0.98inHetaerinacruentatareportedbyCORDOBA-AGUILAR (1994) was derived (in an unexplained way) from greatly fluctuating individualvalues (including nonsensevaluesof > 1).Arecalculation, using datashown infigure 3of TableIV Published estimates ofmeandailysurvivalratesin Calopteryx,HetaerinaandMnaisspecies SSppeecciieess CCoouunnttrryy SSuurrvviivvaall SSeexx RReeffeerreennccee ((ppeerrddaayy)) CC..aaeeqquuaabbiilliiss CCaannaaddaa 00..7777 S3 CCOONNRRAADD&&HHEERRMMAANN((11999900)) 00..7700 59 CC..aattrraattaa JJaappaann 00..9911** <â3 MM11YYAAIKCAAWWAA((11998822)) 00..9933** $$ CC..hhaaeemmoorrrrhhooiiddaalliiss SSppaaiinn 00..9944 <33 pprreesseennttssttuuddyy 00..9944 9$ CC..jjaappoonniiccaa JJaappaann 00..8822//00..8877 6<3 WWAATTAANNAABBEEeettaall..((11999988)) 00..9933//00..8877 59 CC.. vviirrggoojjaappoonniiccaa JJaappaann 00..9944** 6c3 MMIIYYAAKKAAWWAA((11998822)) 00..9977** 99 HH..ccrruueennttaattaa MMeexxiiccoo ((00..9988))//00..9944**** <33 CCOORRDDOOBBAA--AAGGUUIILLAARR(1(1999944)) MM..pprruuiinnoossaa JJaappaann 00..9944 6<3 HHIIGGAASSHHII((11997766)) MM..pprruuiinnoossaa JJaappaann 00..995555 99 NNOOMMAAKKUUCCHHIIeettaall..((11998888)) eessaakkiiii 00..994444 S<3 ssttrriiggaattaa 00..994477 S<3 * Ascalculatedfromhistogramsofsurvivalrates shownin MIYAKAWA(1982),includingonly survival between 1to 10daysaftermarking. ** Recalculatedvalue,seetext. 18 J.J.Beukema CORDOBA-AGUILAR (1994),yieldedanestimateof0.94,andthisfigureismorein linewithrecordsformean dailysurvivalinCalopterygidae by otherauthors.Exclusion oftheabove questionedextreme valuesnarrows therangeofpublished valuesto0.82- -0.97,with apreponderance ofsurvival rates close to 0.94per day (Tab.IV). This may,therefore,beatypical valueforcalopterygid species. Survival duringthefirst dayaftermarking appearstobegenerally lowerthanmean survival, as can be deduced from detailed data given by HIGASHI (1976), NOMAKUCHI etal.(1988)andownobservations(Fig. 1).Itis questionablewhether the enhanced disappearance rates immediately after marking point to higher death rates (by rough handling) or to handling-induced changes in behaviour(stronger tendency to migrate). Theabovefiguresformean dailysurvival inCalopterygidae are mostlyhigher than the values generally recorded in Zygoptera, ranging from0.73 to 0.94 (CORBET, 1999:301). The relatively high valuesfoundinCalopterygidae maypartlybecaused by their site fidelity, conspicuousness, and tameness. This is especially true for reproductive malesandmay(partly) explainthedifferentlevelsformalesandfemales inpercentagesfoundaliveasshowninFigure la.Probably, therewasnorealdifference in dailysurvival ratesbetweenmalesandfemalesinthestudiedpopulation, since the datasetsofthetwo sexes showninFigure larun parallel. Differencesbetweenmales andfemaleswerealsosmallinotherspecies (Tab.IV)andwerenotconsistently higher ineithersex. Theeffecton estimatesofsurvivalrates ofoverlooking a certainproportion ofthe marked individuals can be judged to some extentby omitting all observations on individualsthatweremarkedshortly beforetheendoftheobservationperiod ineach year. Ifthecalculationsare limitedtoindividualsthatcouldpotentially beresighted for>7days, thepercentagesresighted slightlyrosefrom84to86inmaturemalesand from78to 82inmaturefemales.Notethattheextensionoftheobservationperiod to >7 daysappearstohaveenhancedthechancetore-encounter afemalemorethanthat ofamale,as expected fromdifferencesintheirconspicuousness (particularly during ovipositionfemalesmayeasilybeoverlooked). Thoughallestimatesofsurvivalrates fromresightings areboundtobeunderestimates,the biasappearstohavebeensmall inthepresent study. Ameandailysurvivalrateof0.94,asobservedinthepresent studyforthetwosexes andagegroups, wouldmean thatthe sizeof the adultpopulation halvesinabout 11 days. This is roughly the durationofthe maturation period in C. haemorrhoidalis (HEYMER,1973;CORDERO.1989;ownobservations). Maximumlongevity cannot bestatedexactly, because the numberof very oldindividualsthatcan beobserved depends amongotherthings on the population size and the searching effort of the observer (BEUKEMA, 1989). Atthe abovesurvival rate, about 1% ofallemerged individualswillliveatleast74days as anadult.Inapopulationofthepresentsize, the maximumadultlifespanwouldthusamounttotwo orthreemonths(ifthere wereno increasedmortality at highage by senescence). In the field, the oldestindividuals