Reproduction in the Short-beaked Echidna, Tachyglossus aculeatus: Field Observations at an Elevated Site in South-east Queensland Lyna. Beard and Gordon C. Grigg Department ofZoology and Entomology, The University ofQueensland, Queensland 4072,Australia. Beard, L.A. and Grigg, G.C. (2000). Reproduction in the short-beaked echidna, Tachyglossus aculeatus: Field observations at an elevated site in south-east Queensland. ProceedingsoftheLinneanSocietyofNewSouth Wales 122, 89-99. Aspartofaradiotelemetric study ofechidnas (Tachyglossusaculeatus) in south-east Queensland focussing on thermal relations, we were able to confirm and extend present knowledgeofechidnareproduction.MatingwasconcentratedinJulyandAugust,aselsewhere, butwefoundthatechidnashavetheabilitytoconceivesuccessfullyasecondtimewithinthe oneseason,apparentlyinresponsetolosingthefirstyoung.Echidnasinthisareaofsouth-east Queenslandmaybeabletoattemptbreedingeveryyear.Ourdatasupportspublishedestimates ofgestation in the range of20 to 23 days. Females spent two to three weeks in a plugged 'incubation'burrow,maintainingahighandstablebodytemperatureforaperiodencompassing thelastfewdaysofgestation, all ofincubationandthefirstfewdaysofthehatchling's life. Thesingleyoungwascarriedinthefemale'spouchfor45-50days,attainingabodyweightof approximately200gbeforebeingstowedinadifferentplugged'nursery'burrow.Wedescribe thefirstdetailedtimingofafemale'svisitstosuckleheryoung. She visitedregularly, every sixdaysatfirst, graduallyincreasinginfrequency toabouteveryfourdaysbefore the visits ceasedand,presumably,thenewly-independentyoungemergedatacalculatedfiveandahalf monthsofage. Manuscriptreceived 16August2000,acceptedforpublication22November2000. KEYWORDS: echidna,monotreme,radiotelemetry,reproduction, Tachyglossusaculeatus. INTRODUCTION . Short-beakedechidnas (Tachyglossusaculeatus)areextremelycrypticinthewild and rarely breed in captivity. This makes studies oftheir reproductive biology difficult andpiecemeal. Early information was gained from dissection ofdead specimens, single opportunistic observationsonanimals inthewildandsequentialobservations offemales and their young taken into captivity. This identified the breeding season (July-August) andtheegg incubationperiod(10.5 days) andprovidedadetailedpictureoftheanatomy and biochemistry of the reproductive organs and lactation (see Griffiths 1968, 1978). However, comprehensive information on life history parameters and behaviours was lacking. The advent of radiotelemetry made it possible to obtain information about reproduction by echidnas in their natural habitat. However, early telemetric studies concentratedonotheraspectsofechidnabiologysuchashomerange(Augeeetal. 1975). Some short-term telemetric studies by Green et al. (1985) and Griffiths et al. (1988) Proc.Linn.Soc.n.s.w., 122.2000 90 REPRODUCTIONINECHIDNAS addedinformationonmilkintakebyechidnaoffspringandsucklingbehaviouroflactating females, as did an observation on a single female echidna by Abensperg-Traun (1989). Long-term radiotelemetry studies ofechidnas in the wild began in 1986, focussing on thermal relations (Grigg et al. 1989, 1992) and also providing some ofthe first longer- term field observations on echidna reproduction (Beard et al. 1992). Shortly thereafter, Rismiller, working on Kangaroo Island offthe southern coast ofAustralia, commenced anextensive long-term study onechidnas in the wild, concentratingonreproduction and behaviour(RismillerandSeymour 1991; Rismiller 1992, 1999;RismillerandMcKelvey 2000). Furtherfieldworkwhichmay contribute toknowledge ofechidnareproduction is currentlyunderwayalsoinTasmania(S. NicolandN.Andersen,UniversityofTasmania, pers. comm.) While anumberofthe questions aboutechidnabreeding havenowbeen answered, thereis stilluncertainty aboutmany aspects, especiallythose dependantondirectand/or longtermobservation. Ourcontinuingradiotelemetric studyofechidnathermalrelations insouth-easternQueenslandhas providedan opportunitytomake suchobservations and tocompareandcontrastthemwithwhatisknownofechidnareproductionandbehaviour in more southern areas ofAustralia. MATERIALS AND METHODS The study site comprises parts of several grazing properties between Texas and Stanthorpe in SE Queensland at an elevation of 500-1000 metres, centred on 28°4rS, 151°32'E. The area is a mixture of mostly cleared, undulating grazing paddocks with varying grass cover depending on the season, and scrubby, mostly uncleared gullies. Echidnaswerecapturedopportunisticallyand,withclearancefromtheUniversity'sanimal experimentation andethics committee, were implanted with temperature-sensitiveradio transmitters (Austec Enterprises, Canada and Sirtrack Ltd., New Zealand). The transmitters, fittedwith an internal loop antenna, werecoatedin abiologically inertwax mixture ('Elvax' / paraffin wax 20%/80% w/w) and implanted in the peritoneal cavity. They servedtolocateechidnasbyradiotrackingaswellastotelemeterbodytemperature. Signalswereacquiredusingavehicle-mountedomni-directionalwhipantennaandTR-2 receiver -i- TS-1 scanner (Telonics, USA) and/or tracked on foot with a hand-held H- antenna (Telonics, USA) and receiver. One female also had, in two consecutive years, a waterproof, wax-coated temperature-sensitivedatalogger('Tidbit',OnsetCorp.,USA)implantedintheperitoneal cavity. Combined weight ofthe transmitterplus datalogger was approximately 50gm in ananimalofaverageweight3.25 kg. Bodytemperaturedatafromanimalswithimplanted transmitters could often be recorded automatically using a system which consisted ofa timerswitchingon andoffareceiver/scannerandtaperecorderatpresetintervals (Grigg et al. 1990). The same system could be used to monitor time spent by a female in a burrow, using a low gain antenna, or a feedline alone, placed on the ground above the burrow, so that it would pick up a signal from the female only when she was in the burrow orvery close, during entry and exit. Any condition or activity which may have been related to breeding was noted for each animal during thecourseoftrackingandweighingforotherstudies.Asechidnasare normally solitary, aggregations ofmore than one animal wereassumedto signal possible breeding-related activity. The presence of pouch young was obviously evidence of breeding, whilean enlarged, loosepouch was taken toindicateits veryrecentvacationby a burrow-sized young (Griffiths 1968). Other signs included swollen mammary glands which were obvious when the female's belly was exposed and from which droplets of milk could often be extruded, especially under anaesthetic for transmitter implantation. A swollen cloacal area in males was also taken to indicate some reproductive activity because, in this condition, the penis often partly everted when the animal was handledor Proc. Linn. Soc.n.s.w., 122. 2000 L.A.BEARDANDG.C.GRIGG 91 anaesthetised. Normally the penis, when not in use, sits in a fully-enclosed sac offthe cloaca. Weinferreddatesofsuccessfulmatingsbyworkingbackwardsfromtheestimation ofthe ages ofpouch young using growth curves fromGriffiths (1978, in litt.) and Green et al. (1985) and/or time in a brooding burrow, assuming an incubation period of 10.5 days (Griffiths 1978) and a gestation period of20-23 days. We use the term gestation to describe that period between fertilisation and when the egg is laid. Fertilisation was assumedto occur shortly aftermating (see Rismillerand McKelvey 2000). Pouchyoungweresometimesremovedfromthepouchforweighing,oftenwiththe female anaesthetised lightly withhalothane until sherelaxedenough forthe young to be removed without a struggle, weighed and then replaced. The young were apparently insulated enough by the pouch to escape the effects ofthe anaesthetic. RESULTS Twenty-oneechidnas,fromatotalof30animals(20malesand 10females)which werecapturedinthestudyareaovernineyears,providedobservationsrelatedtobreeding. Individualsremainedinthestudypopulationforlengthsoftimevaryingfromafewmonths to several years. Those we lostpresumably hadtransmitters fail oremigrated (Table 1). Table 1. Echidnasincludedinthisstudy. BW Echidna# Sex Captured Lost atcapture Timefollowed (kg) (months) 52 F 26.4.90 19.1.91 4.45 9 M 56 24.5.90 17.3.91 4.2 10 M 63 12.7.90 7.4.93 3.75 33 65 F 27.7.90 20.2.91 3.8 7 M 66 27.7.90 21.10.90 3.9 3 M 67 2.8.90 19.12.90 3.8 4.5 23.10.92 11.10.93 4.0 11.5 68 F 6.9.90 22.8.92 2.85 23.5 69 F 11.10.90 11.10.93 2.95 36 M 77 19.6.91 11.10.93 1.45 28 M 81 30.8.91 29.10.92 3.25 14 83 F 28.1.92 21.7.93 3.45 18 M 84 7.9.92 23.7.94 ?atcapture 22.5 4.3 (10.2.92) 85 F 22.9.92 6.4.93 2.55 6.5 M 87 29.10.92 11.10.93 7 11.5 90 F 20.7.94 present 3.35 54+ M 91 20.7.94 2.9.95 3.8 13.5 M 92 9.7.95 present 5.2 42+ 93 F 23.11.95 present 3.7 38+ 94 F 30.11.95 present 3.55 38+ M 95 19.3.96 present 4.35 35+ M 100 9.10.96 present 4.4 27+ Proc.Linn. Soc.n.s.w., 122.2000 92 REPRODUCTIONINECHIDNAS Mating and its timing. Mating activity in this area is focussed in the second half of July and in August (Fig. 1). However, we saw males keeping company with females for longer than this (Fig. 1), sothepotentialformatingextendsoutsidethis period, as inthe caseofafemale which, in October, replaced an egg lost earlier (see below). This second mating must haveoccurred on orabout22 October, making itthelatestmating yetrecorded. If, as we haveassumed, aswollencloacais aguide, males are availableandreadybothbeforeand afterthe season in which mostmatings arefocussed(Fig. 1). Mostcommonly, mating in this areaoccurredtwo to three weeks afteremergence fromhibernation. However, some femalesmatedwithinaweekofemergence,aswastypicalintheKosciuskostudy (Beard et al. 1992). In one instance we observed amale and female (#52) together less than six days after the female had emerged from hibernation (Fig. 2). Abrasions noticed around thecloacaofthefemaleafterthisencountersuggestedthatthiswasamatingandsubsequent observations of the size of the young indicated that it was successful. We have never found more than two males with a female at the same time. Nor did we see evidence of thematingrutsortrenchesdescribedonKangarooIsland(RismillerandSeymour 1991). Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec A k X X X Figure 1.Seasonaloccurrenceofmalesbeingfoundwithanenlargedcloacalarea/penis(X),beingfoundwith afemale(triangle),andtimesatwhichsuccessfulmatingsoccurred,asinferredfromsizesofyoungandother indirectevidence (bardefinestheperiodinwhichthematingmusthaveoccurred). Asecond breeding eventin the same season. The observations presented in Fig. 1 which were made in October were parts ofa secondreproductive event in the same season forthe female involved. Echidna#93 was tracked to a plugged burrow on 17 and 18 September (Fig. 2) and when captured on 25 September had the remains of an egg in her pouch, but no young. Subsequently, on 9 October, she was found in a log with two males. One of the males was not a marked animal and so was taken for transmitter implantation. The other was relocated with the samefemalethenextday inadifferentlogpileandfoundwiththe samefemaleyetagain onournextfieldtripon 22Octoberinarabbitwarren.Weassumetheyhadbeentogether for the intervening period. The discovery of a 72g young in the female's pouch on 17 Decemberconfirmedthatasuccessful matinghadoccurredonorabout22 October. If,as reported by Rismiller and McKelvey (2000), eggshell remains in the pouch no longer than 48 hours after hatching, the period between loss ofthe young and the next mating would have been approximately one month. Gestation period. TTie72gyoungfrom#93,referredtoabove,wasestimatedtobe25-27daysold,witha hatching date of 20-22 November from an egg laid 10-12 November. Assuming a fixed incubationperiodof10.5days(Griffiths 1978),thegestationperiodwasoftheorderof20-23 days.This wasourmostcomplete seriesofobservationsrelatingtogestation period. Another estimate of length of gestation was made, for #90, using data on body temperature (Tb), instead ofobservations ofpairing, to inferthe dateofmating. Beardet Proc. Linn. Srx;. n.s.w.. 122.2(J00 L.A.BEARDANDG.C. GRIGG 93 May Jun Jul Aug Sep Oct Nov Dec Jan #94(1996) H__ ._H B B y Y Y #94(1995) L #93(1996) H_- H BE * MM M A BB B Y E — — #93(1995) L #90(1997) ; : HA y Y E #90(1996) H_ H BB y E #90(1995) , _H YE #90(1994) M no#(1994) Y #68(1991) M #68(1990) M #69(1993) Y #69(1992) H_ _H B B y Y F #69(1990) Y #65 (1990) M B E #52(1990) H__ HMA A B y Y * Remains ofegg found in poucli Figure 2. Seasonal occurrence ofreproductive events associated with individual females, year by year. H- H=hibernation,A= active, unaccompanied, M =foundwith atleastone male, B =underground in an incubationburroworlog,y=carryingsmallpouchyoung,Y=carryinglargepouchyoung,E=emptypouch, L=lactating. al. (1992)reportedthatthebodytemperatures ofpairedechidnas inaretreatwerehigher thanthoseofsohtaryechidnasinsimilarretreatsandattributedthistotheactivityassociated withmating. #90's temperature record shows aperiodofapproximately 30hours during whichbodytemperaturewasmaintainedessentially stableataround30°C (Fig. 3). Ifthis coincidedwithmating, andiflayingdate is inferredinthe sameway asfor#93, i.e. from estimating the age ofyoung, a gestation period of21-22 days is deduced. Anothersetofobservations, less complete, is at leastconsistentwith a20-23 day gestation period. #52 was found with a male and, as evidenced by cloacal abrasions, assumed to have been mated on 2August. She was found again on 21 Septemberwith a pouch young barely able to be held within the pouch and clearly due to be planted in a nursery burrow. The pouch young was not weighed. A gestation period of 20-23 days would have made this young 47-50 days old, quite consistent with our estimates ofthe planting ofyoung in aburrow at45-50 days (see below). Pouch life. Females in this study were found to carry pouch young until 45-50 days after hatching, attainingaweightofapproximately 200g, afterwhichthey weredepositedina burrow.Youngwhich,fromsubsequentobservations, werejustabouttobedepositedina burrow, did not yet have erupted spines but were unable to fit wholly within the pouch. Thefemalesintheseinstanceswereobviouslyexperiencingdifficultykeepingtheirbelly areaandtheyoung fromdraggingonthe groundandwere unabletocurlcompletelyinto their defensive posture, leavingpartofthe young's body exposed. Proc.Linn.Soc.n.s.w., 122.2000 94 REPRODUCTIONINECHIDNAS Incubation burrows. Asfaraswecanascertain,allfemalesoccupiedanincubationretreatcontinuously for a period oftwo to three weeks roughly corresponding to the timejust prior to egg- laying, through incubation ofthe egg and into thefirst weekorso ofthe hatchling's life. Theseincubationretreats,usuallyabandonedrabbitburrowsexceptforonecaseinwhich a log was used, were always 'plugged' with a mound of earth, in a similar way to the nursery burrows in which the largeryoung were later stowed. This was in contrast with overnightsheltersorhibernationburrowswhichwerealwaysunplugged. Pluggingcould be used reliably as an indicator of whether or not an incubating female or, later in the year, aburrow young was in residence as burrows were always left unpluggedwhen the occupant(s) had moved on. While burrows and other shelters are known to be re-used commonly in this area forovernight and shorttermretreat (Wilkinson et al. 1998), we have so far seen no sign ofburrowre-use forreproductioneitherinthe sameyearorin subsequentyears. Female #90didnotusethe same nurseryburrow inconsecutiveyears.Anotherfemaleforwhich wecanbecertain (#93), usedadifferentnurseryburrow forcachingthe youngfromthat used for incubation. She also used a different burrow for each of the two incubation periods she spent in the same breeding season. Body temperature and behaviour within the incubation burrow. During the period spent underground in an incubation burrow, females displayed patterns ofbody temperature which were uncharacteristically high and stable compared to an echidna's normal heterothermic rhythm during the active season. Readouts from theimplanteddataloggerinfemale#90showedamuchreduceddailyrangecorresponding withtheperiodaroundeggincubation,withbodytemperaturesbeingmaintainedrelatively stable at 31-32°C (Fig. 3). Transmitter data recorded automatically from female #93 showed a similar pattern. Although her body temperature varied a bit more, it did not dropbelow29-30°Cwhileshewasundergroundinapluggedburrow.Thiswasincontrast to the moretypical daily sinusoidal pattern ofbody temperatures recordedthe following monthwhen shewasobservedtobeactiveandcarryingherpouchyoung.Athirdfemale, #94, wastrackedon several consecutive occasions overmorethan aweektoaburrow in arubbishdump. Herbodytemperature, althoughnotmonitoredcontinuously,washigher than 32°C whenevermeasured during this time. Twelve days lateron ournextfield trip, we found her out ofthe burrow, with a small pouch young, confirming that she would have been incubating an egg or a very small hatchling during hertime in the burrow. Visits by the female to the nursery burrow. In onecase, we were ableto continuouslymonitormaternal visits to ayounginthe burrow. Female #90 was observed still carrying the young on 16 October. Three days later, on 19 October, she was tracked to a plugged burrow which was subsequently confirmed as a nursery burrow when our automatic data recording system monitored repeated visits by the female. The first two visits lasted approximately 24 and 17 hours respectivelyandwereseparatedbythreedays.Thereafter,thefemalevisitedatremarkably regular intervals, starting about six days apart until early December and gradually increasing frequency to every five then every four days until the burrow was no longer visited after 13 February. In contrast to the longer visits in the first few days, none of these subsequent visits exceeded five or six hours and most were around three to four hours with atendencytowards even shortervisitsoftwotothreehours in the secondhalf ofthe young's burrow life. When compared to the normal spread ofbody temperatures for this female at this time of year, her body temperatures while visiting the nursery burrow were consistently in the higher 'active' range, above 30"C. The visits happened most often during the pre-dawn to mid morning hours, after the female's usual active period, although they were not restricted to these times. Unfortunately equipment failure prevented us from determining the exact end of Proc.Linn. Soc.n.s.w., 122. 20(XJ L.A. BEARDANDG.C.GRIGG 95 26-Mar 26-Apr 27-May 27-Jun 28-Jul 28-Aug 28-Sep 29-Oct 29-Nov 30-Dec 30-Jan 11-Nov 12-Dec 12-Jan 12-Feb 15-Mar 15-Apr 16-May 16-Jun 17-Jul 17-Aug 17-Sep Figure3.Bodytemperaturessampledhourlyusinganimplanteddataloggerfortwo 11 monthperiods,from March 1996toNovember 1996(above)andNovember 1996-October 1997(below)forfemaleechidna#90, showing the heterothermy associated with the active seasons, two winterhibernation seasons and, particu- larly,thereducedheterothermyassociatedwithmating(singlearrow,M)andwithincubation(joinedpairof arrows).T=torpor,H=hibernation,A=arousal. Proc.Linn.Soc.n.s.w., 122.2000 1 11 96 REPRODUCTIONINECHIDNAS burrow life, but we know it was between 29 January and 13February. That the burrow hadbeenvacatedwasconfirmedbyitbeingleftunpluggedcoincidingwiththeautomatic readings showing no more visits by the female, and the appearance ofanother possible exithole. These observations indicate thatthe young spent aboutthree and ahalfto four months inthe burrow, whichwouldmakeitapproximatelyfivetofiveandahalfmonths old at independence. As we did not excavate the burrow, because we were interested to see whether the female would reuse it the following year, it is possible that the young died or was killed prior to emergence. The female did become pregnant again the next yearbut didnot re-use thatburrow. Breedingfrequency. Every female that we followed formore than one breeding season showed at least some evidence ofbreeding-related activity in consecutive years (Table 2), although we do not know whether they were successful in conceiving and rearing young to independence in all cases. Sudden weight gains exhibited by a couple ofanimals some time afterdepositing theiryoung in anursery burrow may indicate loss oftheirsuckling because, as Rismiller (1999) also observed, female echidnas maintained or slightly increasedweightonce theiryoung were deposited in aburrow. Sexual maturity and body size. Wefoundevidencethatgrowthcontinuespastsexualmaturity.Female#90,which exhibitedbreedingactivity inthreeconsecutiveyears, continuedtogrowthroughoutthat time, at approximately 4.8 mm/year, while her weight varied up and down between 3.0 kg and4.05 kg. This supports RismillerandMcKelvey's (2000) opinionthat size and/or weight is probably not a good indicatorofmaturity. Table 2.Evidenceofbreedinginconsecutiveyears Echidna# Year Details 68 1990 Attractedmalefollower,6/90 199 Attractedmalefollower, 8/9 69 1990 I60gpouchyoungdiedduringfirstcapture 1991 noyoung-hibernateduntil2 19 1992 Young rearedsuccessfullyatleasttoburrowage-nfi* 1993 Young rearedsuccessfullyuntildeathofmotherjustpriortoburrowage 90 1994 Initiallycapturedincompanyofamale,July-nfi 1995 Young apparentlysuccessfullyreareduntilindependenceinFeb96 1996 Young rearedsuccessfullyuntilatleast40days-nfi 1997 Young rearedsuccessfullyatleastto 145g(38-40d?)-nfi 1998 Attractedmalefollower, noyoungobserved 93 1995 LactatingwhencapturedinNov., trackedtosuspectednurseryburrow-nfi 1996 r'pregnancylostyoungateggstage 2"**pregnancyyoungrearedsuccessfullyatleastto72gandprobablytoburrow age 1997 Foundwithmale9/97 1998 Foundwithmale-subsequendysuspectedyounginburrowbutnotsighted 94 1995 LactatingwhencapturedinNov. 1996 Young rearedsuccessfullyatleastto 1lOg 1997 Found<2mfrommale9/97 nfi=nofurtherinformation Proc. Linn.Soc.n.s.w., 122.2000 L.A.BEARDANDG.C.GRIGG 97 DISCUSSION The most significant observation reported here is the second breeding event by a female in the same season. It was apparently stimulated by the loss ofan earlieryoung. RismillerandSeymour(1991) notedthatitwas unknown whetherafemale whohas lost her egg or pouch young early in the breeding season could breed again. Our single unequivocal observation shows that they can, but how commonly remains unknown. A reportofaneggseenon22October(Griffiths 1978)maybeanother,hithertounrecognised exampleofthesamething.ThisphenomenonmaybemorelikelytooccurinQueensland thanincoolerpartsofAustraliabecauseofthelongertimeavailableforyoungtodevelop andforthefemaletoregainconditionbefore theonsetofacomparatively shorter, milder winter.Theonemonthperiodobservedbetweenlossofthefirsthatchlingandthe second mating may be relevant to interpretation offuture data on echidnaoestrous cycles. Ourstudy, althoughreporting observations fromone animal only, has providedthe mostcompletepicturesofarofmaternalcarefollowingcachingoftheyounginaburrow. Griffiths et al. (1988) documented suckling visits at intervals offive to ten days by one female,usingdisturbanceofsticksplacedacrosstheburrowentranceasaguide.However, their observations started only when the young was approximately 150 days old. Abensperg-Traun (1989) also made afew observations on one animal, with the shortest noted interval between visits being three days immediately following the caching ofthe young in the burrow. Rismiller and McKelvey (2000) reported suckling for two hours every five to six days, although there is no information provided on the number of observations orthe age ofthe young. Theyoungprobablygoestorpidatleastsomeofthetimebetweenfeedingvisitsby themother, asobservedinoneinstancebyGriffithsetal. (1988).Thefemalemayneedto warm up her infant before feeding, which is consistent with the higher 'active' body temperaturesobservedforthemotherduringfeedingvisits. Indeed, itisnotyetknownat whatstagetheburrowyoungacquirestheabilitytothermoregulate.Alsonoteworthywas therelatively short lengthofeach visit. Echidnas areknown tohave aprodigious ability to imbibe large amounts ofmilk in a very short time (Green et al. 1985; Griffiths 1989) andfromayoungage. Greenetal. andGriffithsreportedthatevenyounginthepouchgo threetofourdayswithoutfeeding.Theresultspresentedhereareentirelyconsistentwith this picture. Anothersignificantobservationreportedhere is the furtherconfirmation oftighter body temperature (Tb) regulation by the female while underground around the time of egg incubation, firstdescribedforfemales in the Kosciusko area (Beardet al. 1992) and confirmedby Nicol andAndersen (2000) inTasmania. It seems likely thatthe higherTb offemalesinanincubationburrowmaybeassociatedwithmuscularactivity. Frequently, when animals were tracked to 'plugged' incubation burrows, the signal strength was variable, whichwouldusuallybetakenas anindicationofmovement(Griggetal. 1992). As these incubating females were obviously not going anywhere, there is a strong possibility that they were moving or shivering to generate the heat required to maintain the higher and more stable than normal body temperatures observed. Thistighterbodytemperatureregulationaroundthetimeofincubationisincontrast tothetypicalechidnaheterothermicpatternwhichwasobservedduringmostofgestation, but the significance of this is unclear. Perhaps a constant, warmer temperature at this stage is necessary to maintain development. It is hard to see, however, why there is a premium on temperature regulation during incubation of the egg and not throughout gestation. Clearly, furtherwork is needed. The breeding season in our study area is broadly similar to that found in other places in Australia (Griffiths 1989; Beard et al. 1992; Abensperg-Traun 1989). As emergence from hibernation presumably sets the earliest date for production ofyoung, minorvariationinthetimingofbreedingmayberelatedtoregionaldifferencesintemporal hibernation patterns. Proc.Linn.Soc.n.s.w., 122.2000 98 REPRODUCTIONINECHIDNAS OurestimatesofgestationperiodagreewellwithRismillerandMcKelvey's (2000) estimateof22 to 24days. Itshouldbenoted, however, thatthere is noreason to suppose that the timing of these events is as fixed as typical data from more homeothermic endotherms.Fig.3 showshowvariableafemaleechidna'sbodytemperaturecanbeduring the gestation period. They may evenbecome torpid, as reportedby Geiser and Seymour (1989). Assuming developmental rate is temperature dependant, this pronounced heterothermy mayhelpto explain whytherehavebeen suchvariableestimatesreported, including9-27days (Griffiths, 1989), 18-27days (Rismiller 1992),22-24days(Rismiller and McKelvey 2000), 19-28 days (Knee 1998) and 20-23 days in this study. The carriage of young in the pouch, reported here, has been noted on Kangaroo Island, around Canberra and in Western Australia. However, in the Kosciusko area we saw no young being carried (Beardetal. 1992). Presumablythey were leftintheburrow once the female started foraging again after the birth, so this feature of reproductive behaviourmay vary geographically. Ourestimate in south-east Queensland of45-50 days carriage in the pouch before young are deposited in a nursery burrow is similar to the average of 53 days given by Griffith (1989), the estimate of50-55 days quoted by Abensperg-Traun (1989) and the range 45-55 days given by Rismiller and McKelvey (2000). It is likely that the end of pouchlifeisdictatedbytheyoungreachingasizewhichistoobigforthefemaletocarry comfortably and safely. We foundthatthe young weighedabout 200g when depositedin anurseryburrow, in agreement with the lower end of Rismiller and McKelvey's (2000) observed body weights of 180-260g. Griffiths (1978) observed young still being carried in the pouch at 260g. This difference could be a consequence ofQueensland echidnas being generally smallerthanthose in some southern mainland areas andhaving correspondingly smaller young. Green et al.'s (1985) data suggest that, after they reach 90g, growth rate ofthe young echidna is proportional to the mother's weight. Rismiller and McKelvey (2000) also supported the speculation that smaller young are producedby smaller mothers and that the size ofthe mother may also influence time to weaning. This may help explain why our estimate ofage at weaning, 5-5 V2 months, is somewhat younger than the 6-7 months reported by Rismiller and McKelvey (2000) and suggested by Griffiths et al.'s (1988) observations. The smallest breeding female we encountered weighed 2.9 kgjust afterher young hatched. Theuseofpluggedincubationburrowshasbeenreportedinotherstudies.Abensperg- Traun(1989), workingintheWesternAustralianwheatbelt,recordedanunbrokenperiod ofatleasteightdays spentundergroundbyafemaleechidnaaboutthetimeofincubation andhatchingofheryoung. Rismiller(1992)andRismillerandMcKelvey(2000)observed the use ofan underground retreatby females forupto ten days during egg incubationon Kangaroo Island, but noted that it is facultative, with many females being found out foragingduringthistime. Incontrast,ourworkinAustralia'sSnowyMountainsdescribed anunbrokenperiodoffourtosixweekscoincidingwitheggincubationandthefirstthree weeksormoreofthehatchling'slife, withtheyoungneverbeingobservedinthefemale's pouch. This pattern has also been reported for Tasmania (Knee 1998). The practice of 'plugging'orback-fillingnurseryburrowsforolderoffspring, whichseemstobeafeature ofall areas, may serve to discouragepredation or, as suggestedby Griffiths etal. (1988), to maintain 'equable' conditions within the burrow. Several ofourobservationsarein contrasttostudieson KangarooIsland(Rismiller and Seymour 1991; Rismiller 1992) and may be related to differing echidna densities. Rismillerand McKelvey (2000) recordedapproximately nineanimalsperhectare, which is almost certainly higher than many other places in Australia, and certainly higher than in ourstudy area. Males on Kangaroo Island generally follow females forlongerperiods (uptoseven weeks)before matingoccurs, often forming 'trains'ofupto 1 1 animals. Our matingaggregations apparently occurred forshorterperiodsandinvolvedfeweranimals, which may reflect decreased competition due to lowerdensities. The lack ofmating ruts Prcx:.Li.sn.Soc..n.s.w., 122. 2000