Day&Hyphen; and Night&Hyphen;Time Movements of Radiomarked Red Knots Staging in the Western Wadden Sea in July&Ndash
Total Page:16
File Type:pdf, Size:1020Kb
36 Day- and night-time movementsof radiomarkedRed Knots stagingin the westernWadden Sea in July-August 1995 Jan van Gils & Theunis Piersma Van Gils, J. & Piersma,T. 1999. Day- andnight-time movementsof radio-markedRed Knotsstaging in the westernWadden Sea in July-August 1995. WaderStudy Group Bull. 89:36 - 44. Fourteenindividually radio-marked Red KnotsCalidris canutusstaging in the westernWadden Sea in July-August1995 oftenwandered considerable distances in the courseof a singletidal cycle,which confirmedprevious ideas based on movementsof flocksobserved. The birdscontinued this tidal feedingrhythm during nighttime darkness. There was a tendencyfor Red Knotsto remainfor shorter periodson thehigh-tide roost at nightthan during the day. Individual birds with bodymasses > 140 g when radio-marked(which may havebeen close to departurefor WestAfrica), were lessitinerant than light-weight individualsand were more likely to feed in an area closeto the roostwhere relatively soil-shelled shorecrabsmade up muchof the diet. Suchmass-related differences in foraging movementsmay be explainedby migration-relatedphysiological changes, different needsto track changesin local food stocksand to find suitableflock-mates to take-off with andbody mass related predationand starvationrisks and flight costs. Jan van Gils & TheunisPiersma, Netherlands Institute for Sea Research(NIOZ), P.O. Box 59, 1790 AB Den Burg, Texel,The Netherlands, and Centrefor Ecologicaland EvolutionaryStudies, University of Groningen,P.O. Box 14, 9750 AA Haren, The Netherlands INTRODUCTION at the individuallevel duringday and duringnight. We aimed Red Knots Calidris canutusbreed in the High Arctic and to testwhether: (1) individualRed Knotsstaging in the spendthe non-breedingseason on intertidalmudflats in westernWadden Sea in late summer/earlyautumn can be temperateand tropical zones, feeding mainly on buried confirmedto covera large areaof intertidalflats duringsingle bivalveprey. In the westernWadden Sea, The Netherlands, low waterperiods, and (2) whetherthis rangingpattern is where this studytook place, two subspeciesoccur. Members maintainedat night. We haveused radiotelemetry as a means of the islandicasubspecies, breeding in Greenlandand to achieve this aim. For the benefit of readers, references on NortheastCanada, stay t¾om August to April, anduse the area the useof radiotelemetrym studiesof waderbehaviour have as their moulting and wintering site (Davidson& Wilson beenassembled m an Appendix. 1992). The canutussubspecies, breeding in north-central Siberia,only stopsover in July-Augustand in May, on the MATERIAL AND METHODS way to andfrom their West-African winter quarters(Pierstoa At Richel,one of the mostimportant wader roosts m the et al. 1992). westernWadden Sea, we mistnetted52 Red Knots in late July 1995. We glued radio-transmitterson the backsof 14 Red In the westernWadden Sea bothpopulations make intensive Knots,using superglue rather than epoxy (Raim 1978), but use of the totertidal flats around the small uninhabited islands otherwise followed the restructions of Warnock & Wamock of Griend and Richel for feeding. The two islets,which are (1993). Basedon the absenceof primarymoult, 10 of these free of mammalianpredators but not necessarilyof avian RedKnots presumably belonged to the canutussubspecies predators,are usedas roostingsites during high tide. In an breedingin Siberia. The otherfour birdsbelonged to the earlierstudy, Piersma et al. (1993a) providedevidence based islandicasubspecies, one being a one-yearold bird in on movementsof flocks observedin the day-time that Red advancedwing moultthat waslikely to havespent the summer Knotsmay cover an area as large as 800 km2 m a coupleof m theWadden Sea, the restbeing adult birds in wing moult t•dal cyclesm their searchfor food and saferoosting sites. (Table 1). All birdswere molecularly sexed using small blood samplesstored in 70% ethanol (Baker et al. 1999). The birds The large-scalemovements described by Piersmaet al. wereweighed to the nearestgram on an electronicbalance, (1993a) were basedon observationsof flocks duringdaytime. theirbill length,total head, and tarsus plus toe length Here we reporton our first attemptto measurethese patterns measuredto thenearest 0.1 millimetrewith calipers,and wing 37 Richel "/.•- ..•.......... • qemm.. ..- ß •, .'r_•'.. , - . Gn.'end. _ ., ß . Richel Ballastplaat Friesland Richel . ß ,•, Ballastplaat ß ß Figure 1. Examplesof movementsof individualRed Knots through the western Wadden Sea. Filledsquares indicate positions during the mght,open squares are positions in daytime.[A]. Movementpattern of femalecanutus-Knot (#212), 1-4 August.In thefirst night,between 0 00-3:55 h, the birdis roostingat Richel,whereafter she feeds south of Richeluntil 7 h. Duringthe nextlow tideperiod she feeds west of Griendfor a shorttime, beforemoving to the flats eastof Griendin the evening. Next morningshe is foundnortheast of Griendtill noon. That lateevening until the earlymorning she is foundon the mudflatsclose to the Frisiancoast. [B]. Almostthe wholeof 13August was spent by thesame female canutus-Knot #212 at andnear Richel, before moving on to theflats near Vlieland during the night. Duringthe subsequent daytimelow tideperiod she was northeast of Griend,where she was recorded again at nightuntil late morning. That evening she was found to passwestward south of Griend,probably flying from feedinggrounds in the east. Sheroosted at Vlieland, whereaftershe fed on Posthuiswadat night. [C]. Thismap shows the movements of femaleislandica-Knot (#249), over a periodof almost100 hours, 11-16 August.In thefirst mghtthe bird is roostingat Griend,but flew to Richel(a distanceof 9 km) in the earlymorning where she probably continued to roost.Then shestarted to feedat theflats around Richel. At noonthat day, she was again recorded roosting at Richel. The nexthigh tide sheis roostingat Griend. That night sheflew to Richelwhere she continued roosting, to be recordedfeeding close to Richelthe next low tide. About half a day latershe is feedingsouth of Vlieland(7 km away),and at theend of thatday shefeeds east of Griend,implying another flight of 23 km. Next morningshe is foundthere once more, whereafter she roosts at Griendin the afternoon(a distanceof 7 km). That nightshe is recordedsouth of Vlielandagain, presumably feeding (16 km). Overthe period of observationthis bird has flown a minimumof 80 km, or 20 km/day. 38 Table 1. Summaryof biometrics,sex, age and subspeciesof the 14 Red Knotsradio-marked in late July 1995. All birdswere caughtat Richel, duringnight-time high tides. Radio frequencyat the 173 MHz bandwidthis given. (c.y.=calendaryear). Radio Capture Body Bill Total Tarsus Wing Sex Age Putative frequency date mass length Head + toe length sub (kHz) (g) (mm) (mm) (mm) (mm) species 212 27 July 139 35.2 64.6 58 180 female adult islandica 220 27 July 176 34.8 64.4 59 173 female adult canutus 230 28 July 138 35.3 67.4 57 167 female adult canutus 235 28 July 132 33.7 63.4 56 171 female adult canutus • 240 29July 138 33.4 62.0 57 172 male adult islandica 249 29 July 140 32.4 63.2 57 180 female adult canutus 259 29 July 123 34.5 65.8 57 178 female adult canutus 280 29 July 186 36.5 65.0 61 175 female adult canutus 289 29 July 141 35.2 64.2 58 173 female adult canutus 299 29 July 142 35.6 65.3 58 175 female adult canutus 302 29 July 145 35.6 67.2 58 175 female adult canutus 314 29 July 112 32.2 63.6 56 167 male adult canutus 324 29 July 130 30.9 63.6 57 157 female 2nd c.y. islandica i 341 29 July 141 34.1 65.0 59 177 female adult islandica length (maximum chord) was measuredto the nearest Table2 (a).An analysisof varianceof thesignal strength of thefree millimetreusing a stoppedruler (Table 1). transmitterreveals the effects of (1) the heightof the transmitter,(2) the distancebetween receiver and transmitter, (3) theheight of the We used 1.4 g radio-transmittersfor the 173 MHz bandwidth receiver(4) an obstacle(a humanbody) just beforethe transmitter (Holohil SystemsLtd., Carp, Ontario,Canada). Individual and(5) theangle of thetransmitter antenna (tail awayor transmitterfrequencies differed by about 10 kHz. Battery-life perpendicular).Only distancewas treatedas a continuousvariable, time accordingto manufacturerwas 42 days,and was at least all othersas categoricalvariables. All variablescontributed 36 daysin the field. Signalswere listenedto by using significantlyto theexplained variance (p<0.0001; R2=0.60; n=380). hand-held three-element Yagi antennasand portablereceivers (TRX-2000S, Wildlife Materials Inc., Carbondale, Illinois, variance (%) USA). The antennaswere mountedat the end of long poles Variable Explainedp-value transmitterheight 29.1 and positionedapproximately 10 m aboveground level. The <0.0001 distance(km) 22.5 <0.0001 • signalwas detectedby manually turningthe pole 360ø . When a signalwas heard,the antennawas pointedin the direction receiverheight 4.2 <0.0001I giving the strongestsignal. As soonas this directionwas obstacle 2.3 <0.0001i found, backgroundnoise, amplitude of the signalas visible on the displayand directionwere registered.In caseswhere no transmitterangle____ 2.1 <0.0001• Table 2 (b). Parametervalues for the above model. signalwas detected,only the level of backgroundnoise was noted. Variable Parameter Effect constant +0.64 We made a detailedassessment of the signalstrength by transmitterheight 2 m +0.27 varyingthe positionsof both a receiverand a 'free'transmitter transmitterheight knot height -0.03 on a clear and sunnyday. This testrevealed strong effects on transmitterheight