REFERENCES Maynard-Smith,J. 1982. Evolutionand the Theoryof Games. Ebro. Tesina de Licenciatura. Univ. Barcelona.

Abranson,M. 1979. Vigilanceas a factorinfluencing flock formation Maynard-Smith,J. 1982. Evolutionand thetheory of games. amongCurlews Numenius arquata. Ibis 121:213-216. CambridgeUniversity Press. Cambridge. Baker, M.C. & Baker, A.E.M. 1973. Niche relationshipsamong six Metcalfe,N.B. 1989. Flockingpreferences in relationto vigilance speciesof shorebirdson theirwintering and breedingrange. benefitsand agressioncosts in mixed-speciesshorebird EcologicalMonographs 43:193-212. flocks. Oikos 56: 91-98.

Burton,P.J.K. 1974. Feedingand thefeeding apparatus in waders: Paulson,D.R. 1990. Sandpiper-likefeeding in Black-Bellied a studyof anatomyand adaptationsin the Charaddi.British Plovers. Condor92: 245. Museum Natural History. Pienkowski,M.W. 1978. Differencesin habitatrequirements and Cramp.S. & Simmons,K.E.L. 1983. The birdsof the Western distributionpatterns of ploversand sandpipersas investigated Palearctic. Vol.III. Wadersto gulls. OxfordUniv. Press. by studiesof feedingbehaviour. Verh.ore. Ges. Bayern.23: Oxford. 105-124.

Davies, N.B. 1982. Behaviorand competitionfor scarce resources, Puttick,G.M. 1979. Foragingbehaviour and activitybudgets of Pp•363-380. In: King'sCollege Sociobiology Group (eds), CurlewSandpipers. Ardea 67:111 CurrentProblems in Sociobiology.Cambridge University Press.Cambridge. Pyke,G.H., Pulliam,H.R. & Charnov,E.L. 1977. Optimalforaging: a selectivereview of theory and tests. Quart. Rev. Biol.52: Dawkings,R. 1980. Goodstrategy or evolutionarystable strategy? 127-154. Pp. 331-367. In: Barlow,G.W.and Silverberg,S.(eds). Sociobiologybeyond nature. Westview Press. Colorado. Schoener,T.W. 1971. Theoryof the feedingstrategies. Ann. Rev. Ecol. Syst. 11: 369-404. Goss-Custard,J.D. 1969. The winterfeeding ecology of the Redshank( Tringatotanus). Ibis. 111:338-356. Silliman,J., Scott, M., Ills, G. & Alden, S. 1977. Effectof flock size on foragingactivity in winteringsanderlings. Wilson Bulletin Marshall,W. 1981. The feedingbehaviour and ecologyof the 89: 434-438. TurnstoneArenaria interpres on a rockyshore and in captivity.Unpublished Ph.D. Thesis. Universityof Whitfield,D.R. 1990. Individualfeeding specializations of wintering Edinburgh. turnstoneArenaria interpres. J. Animal Ecol. 59:193-211. Martinez-Vilalta,A. 1984. La invernada de limfcolas en el Delta del Waite, T.A. & Grubb,T.C. 1988. Copyingof foraginglocations in mixed-speciesflocks of temperate-decidouswoodland birds:an experimentalstudy. Condor90: 132-140. Census-efficiency for breeding Common Sandpipers Actitis hypoleucos

D.W. Yalden & P.K. Holland

Yalden, D.W. & Holland, P.K. 1993. Census-efficiencyfor breedingCommon Sandpipers Actitishypoleucos. Wader Study Group Bull. 71: 35-38. The relativeefficiency of detectingCommon Sandpipers during each week in the breeding seasonwas estimatedby repeatedobservations of a colour-ringedstudy population. We were mostlikely to detectterritories (78% detection)during the middleof May, but censusefficiency was consistentlygreater than 50% throughoutthe centralnine weeks of the breedingseason. D.W. Yalden,Dept of EnvironmentalBiology, University, Manchester M13 9PL, UK. P.K. Holland,Rennie Court, BrettarghDrive, LancasterLA 1 5BN, UK.

INTRODUCTION chicks,but are mostlyoverlooked in mid-Aprilto mid-May when they are incubating.Common Sandpipers also In designingany censusingprogramme, it is usefulto seemto be rathercryptic during incubation, especially knowhow detectability of the targetspecies varies with duringthe fourthweek of May, thoughto a less extreme time. Thus Reed et aL (1985) showedthat moorland extentthan Golden Plovers. This note is an attemptto waders are less detectableduring the middleof the day quantifythat impression. than in the early morning(pre-09.00) or evening(post - 18.00); intensivestudies can thereforebe arrangedto exploitthese periodswhen the birdsare moreobvious, METHODS and extensive moorland surveysavoid them in order to achieveeven coverage. SimilarlyYalden & Yalden (1991) Our studyarea and the populationdynamics of our study demonstratedthat GoldenPlovers Pluvialis apricaria are populationhave been describedelsewhere (Yalden 1986; best detected in June, when they are guarding hatched He!!and& Ya!den !99!). Between 7 (! 992) and 22 (!977)

35 pairsof CommonSandpipers occupy about 10 km of the 1990. Duringincubation in late May however,survey RiversAshop and Alport,Derbyshire, UK (c 53ø24'N,01ø efficiencywas lower;only 57.5% of territorieswere found 47'W; NGR SK1489). About 75% of the birds are colour- and thisdrop was statisticallysignificant (Wilcoxon paired ringedand can thereforebe recognisedin the field. For sampletest betweenthird and fourthweeks of May, n=6, this paper,survey results for 1987-1992have been T=0, p<0.05). analysed. Even coverageof the whole studyarea, from the fourthweek of Aprilto the secondweek of July,was Detection of individualbirds, rather than territories, shows attemptedon a weeklybasis in these years. The total a morevariable pattern (Figure lB). As withterritories, numberof territoriesoccupied each year was assessedat detectabilityis high in mid-May;assuming that there the end of each season;occupation was assumedif a pair shouldbe two birdsper territory,67.1% and 64.6% of was presentat leastonce, or a singlebird at leasttwice. them were detected in those weeks. Fewer than half of The resultsof each week's surveywere then reassessed them were detectedduring the next three weeks, but a to establishwhether each territoryhad been detectedon secondpeak of detectability,64.2%, occurredin the third each occasion,and whetherboth of the pairor onlyone week of June when manyparents were guardingyoung per birdhad been seen. Surveyswere assignedto chicks(Yalden & Holland 1992). Detectabilityof birds, as weeks,for comparisonacross years, the fourth"week" of of territories,then fell away as failed breedersleft and as each monthbeing variably of 9 or 10 days. youngfledged and then left their breedingterritories with their parents.

RESULTS Why was surveyefficiency so rarely 100%? Detailed recordsfor the best period,the secondweek of May, are Detectabilityof CommonSandpipers was quitegood presentedin Table 1. Of 98 territory-visitsin that period, throughoutthe breedingseason, with over 50% of the 76 (77.6%) were apparentlyoccupied, mostly by bothof territoriesbeing registered,averaged acrossthe years, in the pair. The 22 (22.4%) failuresto detectthe occupants all weeksfrom early May to early July. The best results are the interestingrecords. In two cases, the territories were achieved in mid May; median values (acrossthe six had been occupiedearlier in the seasonbut birdswere years) were 78.1% and 77.5% of territoriesdetected not seen on this or any later date; they are scored as (Figure1), with a rangefrom 62.5% in 1989 to 100% in failures, and the birds might have died, or moved elsewhere. Eightof the territorieswere only occupied Figure1 (a) Percentageof CommonSandpiper Actitis hypoleucos later in the breedingseason - late arrivals. In twelve territoriesdetected in successiveweeks of the breedingseason. cases, however, the birds were seen both before and after Medianvalues and the rangesfor the six breedingseasons 1987- this date, so were genuinelymissed on the occasionin 1992are shown. (b) Similarresults for the percentageof individual question;even in the periodof peak detectability,in a birdsdetected (assuming 2 birdsper territory). familiarstudy area and with observersexperienced in lOO- • A censusingthis species,one must expectto miss 13.6% i (12/88) of the occupiedterritories.

It is interestingto analysethe detailed recordsfor the fourthweek of May in the same way (Table 2). Of 117

-- territory-visitsin this period,69 (58.9%) were detected, butin only 24 were both of the pair seen, a significant• smallerproportion than in the secondweek of May (X'-= L• I.-- 16.37, p<0.001). Where only a singlebird was seen, 19 were males, 20 were females and 6 could not be scored (not seen well enoughto read the rings,or unringedin a territorywhere neitherbird was ringed). The presenceof a singlebird in a territoryimplies that its mate was

1OO- I I I I I I I I I I incubating,and the even numbers of males and females • B stronglysuggests that they sharethis duty equally, as they do guardingof the youngchicks (Yalden & Holland 1992). Of the 48 (41.1%) territoriesnot detectedat this time, 19 laterhad hatchedyoung and another20 were occupiedon lateroccasions, so these 39 territories escapeddetection even though they were (still)occupied. Onlynine territories escaped detection because they were in fact,apparently deserted by thisdate (occupantshad been seen earlier in the season, but were not seen on this or any later date).

4i 1i 2i • 4i 1i 2i 3i 4i 1i 2i APR MAY JUN JUL

36 Table 1. Detailedresults of censusingCommon Sandpiper Actitis hypoleucos territories during the secondweek of May (the "best"censusing week) (n = numberof territoryvisits)ø

Year (n) Detected Missed

pair maleonly female indet.single failed? late arrival missing only 1987 (14) 9 2 1 - 2 1988 (32) 12 3 5 3 - 1 8 1989 (16) 6 3 - 1 1 3 2 1990 (13) 12 1 - - - 1991 (12) 9 1 - - 2 - 1992 (11) 5 2 1 1 2 Total (98) 53 12 7 4 2 8 12 76 (77.6%) 22 (22.4%)

Table 2. Detailedresults of censusingCommon Sandpipers Actiris hypoleucos territories during the fourthweek of May (peak incubationperiod) (n = numberof territory-visits).

Year (n) Detected Missed

pair male female only indet.single failed? present yet to only hatch 1987 (20) 4 2 4 3 2 5 1988 (15) 2 5 4 - 2 2 1989 (25) 7 4 1 - 3 4 6 1990 (26) 5 4 8 1 7 1 i 99i (24) 5 4 2 1 4 5 3 1992 (7) ! - 1 1 2 2 Total (117) 24 19 20 6 9 20 19 69 (58.9%) 48 (41.1%)

37 DISCUSSION fourthweek of May givesa poorerestimate than the weekseither side. Thirdly,if onlya singlevisit is possible, In manyrespects, Common Sandpipers should be easy the secondweek of May shouldprovide a goodpopulation birdsto census. They are tied to easily recognisedand estimate,probably 80% of the truevalue. Alternativelya easilytraversed linear habitats,and defend shortlengths visit in the third week of June should locate 70% of the (150 - 200 m) of these as theirterritories. Duringterritory territories, and a visit then would also reveal information establishmentin earlyMay, they are quiteconspicuous, on breedingsuccess; records should be differentiatedinto often vocal, in defence of their territory,and both birdsin "alarming"adults (guarding chicks), pairs (probably failed) the pair defend it againsttheir neighbours.Even so, 12% and singleterritory occupants (mate stillincubating a late of the territoriesmay be missedby a single-visitcensus. clutch). By late May, when birdsare incubating,one of the pair is usuallydetected in the territory,though 36% of the territoriesstill occupied at thistime may be overlookedby REFERENCES a single-visitcensus. In June, with many of the still- occupiedterritories having parents noisily guarding their Holland,P.K. & Yalden, D.W. 1991. Populationdynamics of chicks,detectability rises again. By this time, however, CommonSandpipers Actiris hypoleucos breeding along an early-failedbreeders are already leavingtheir territories, uplandriver system. BirdStudy 38: 151-159. so that overalldetection rates are beginningto fall off. By Reed, T.M. & Barrett,C., Barrett,J., Hayhow,S. & Minshull,B. the secondweek in July, it is no longerworthwhile to 1985. Diurnalvariability in the detectionof waderson their attemptto censusbreeding Common Sandpipers in the breedinggrounds. BirdStudy 32:71-74. Peak District,and our visitsafter this time are usually Yalden,D.W. 1986. The habitatand activityof Common targettedat checkingthe progressof late (secondclutch) Sandpipersbreeding by uplandrivers. BirdStudy 33:214- breedingattempts. 222.

The mostimportant practical points to be madeare, firstly, Yalden, D.W. & Holland, P.K. 1992. Relative contributionsof that CommonSandpipers should be censusedin termsof CommonSandpiper Actitis hypoleucos parents to guarding theirchicks. Ringing& Migration13: 95-97. territoriesnot birds:a territorymay be revealedeither becausea pair is seen, or by the presenceof a single Yalden,D.W. & Yalden, P.E. 1991. Efficiencyof censusingGolden bird. Secondly,censusing any time betweenearly May Plovers. Wader StudyGroup Bull. 62: 32-36. and the end of June (in Britishlatitudes, at least)should givea reasonableestimate of the population,though the

SAFRING NEWS AVAILABLE ON REQUEST

$afringNews, the biannualjournal of the SouthAfrican BirdRinging Unit (SAFRING), has been publishedsince 1972. Thejournal publishes articles reporting the results of ringingactivities in southernAfrica, ageing and sexing guidesto southernAfrican birds, and bookreviews. Many of these papersrelate to migratorybirds, and therefore theirimmediate interest transcends a purelyimmediate southernAfrican relevance. Papers in SafringNews are includedin Wi/d/ifeReview and similarabstracting services.

Previously,Safring News has had a limitedcirculation, withcopies being sent only to SAFRINGringers and to ringingoffices. In responseto requestsfrom researchers and institutionsin manyparts of the world,it has been decidedto makeSafring News available on subscription. The subscriptionrate for volume23, 1994, has been set at 30 US $ for individualsand 60 US $ for institutions• Ordersfor subscriptionsshould be sentto SAFRING• AvianDemography Unit, Department of Statistical Sciences,University of CapeTown, Rondebosch 7700, South Africa.

38