Internahonal Wader Studies 10. 152-169 Breeding distribution of Dunlin Calidris alpina in Russia E.G. Lappo & P.S.Tomkovich Lappo,E.G. & Tomkovich,P.S. 1998. Breedingdistribution of Dunlin Calidrisalpina in Russia. International Wader Studies 10: 152-169. Breedingrecords and breeding densities available from the literature,museum collections and unpublisheddata are summarised for six subspeciesof Dunlin Calidrisalpina in Russia.Differences in distributionand habitatsfound in subspeciesand populationsare usedto outlinetheir breeding ranges.Each subspecies or populationhas core areas where demities are high and habitats most diverse,but it was not possibleto discoversuch core areas for all subspecies.Maximum densities arenot the samein differentsubspecies, thus centralis never reaches demities that are as high as foundin alpina,sakhalina and kistchinski. Coastal areas are often the mostdensely populated in Europeantundra and in theFar East. Dunlinavoid dense stands of dwarf birchpossibly because largehydrophilous Enchytraeidae worms, their main food, are scarce there. Thisis probablythe reasonfor Dunlin havinga patchydistribution in southerntundra; only isolatedbreeding localities on bogsare known in areaswhere large dwarf-birch tundra belt is present. E.G.Lappo, Department of Biogeography, Institute of Geography,Russian Academy of Sciences, StaromonethnyPer. 29, Moscow109017, Russia. P.S.Tomkovich, Zoological Museum of Moscow State University, Bolshaya Nikitskaya Street 6, Moscow 103009, Russia. Aanno,E. F., TOMKOBW•, Fl.C. t998. FHe3Aosoe pacnpocTpaHeHue qepHo•6uKa Calidris alpina • Po½cuu. International Wader Studies 10:152-169 O6o61ijeltlaAocTyllltlae 1;auTeparypHiax I,ICTOtlHI, IliaX, My3e•mlx IioAAelilJI4HX • HeOHyi•AHKOBaHHbIX•aHHbIX cAyqal/rHe3•OBaHI/H I/HAOTHOCTb rHe3•oBaHI/H AAH IHeCTI/ IIOAB•AOBqepHo3O6•Ka Calfdris alpina B Pocc•. Ha•AeHHtiepa3.•q•n S pacilpOcTpalteH•• Mi•3TaXO6HTaHHH y no•o• • nonyam• •CnOabay•OTCSASSTORO, qTO6bl oIIp•OAHTb HX rHe3AomaeapeaAta. ¾ sa•AoronoAB•a •A• Ka•Aol r•or•y•m• eCTbOr•T•My•fia apeaAa c stacoso•r•AOTHOCTb•O • Ha•6oAee pa3Hoo6pa3HbtSm 6•OTOna•m, HO TaK•e aOHla He yAaAoc], noAB•Aa•m,Tas qTO r•oAB•A ½entra]i$ H•KorAa He AOCT•raeT Taso• StaCOSO• r•AOTHOCT•, sasas aaper•cTp•posaltaAA•Ir•oAB•Aos alpfna, sakhallna Hkistchfnski. I]p•6pe•Hlae paloma qacTo Hau6oneerycTO aaceneHb• a esponelcKol TyHgpe Ua AanbHeMBocTose. tlepHoao6usn ua6era•oT rycTbrX3apocnel epHusa, MOneT 6brrb, r•OTOMy qTO 6Onbmue rugpo•unbHlae qepBlm {Enchytraeidae:OCHOBHOi UXSOpM) TaM pegKu. HMemto nOSTOMy, sepO•ITHO, qepHoao6usu pacr•pocTpaHemaHepaBHoMepHo BIO•HO• TyHApe; uanecTma numb oTAenbmae uaoauposaHHlae MeCTameagosaHua Ha 6OhOTaX, a paloHax, rge acTpeqaeTca TyHgpoabr/i r•osc c macosusmu rycTbI•m aapocna• ep•usa. Introduction Abouthalf the totalbreeding range of Dunlin is in Russia,with six subspeciesdistributed from Beinga circumpolarspecies, Dunlin Calidrisalpina temperateto high arcticareas (Stepanyan 1990), so hasunprecedented geographic variability among onecan expect that generalphenomena can be arcticand subarcticbirds (Cramp & Simmons1983) tracedin thispart of the range.Available published with up to tensubspecies recognised. Knowledge descriptionsof breedingdistribution of Dunlin in of its breedingdistribution is a basicrequirement thisvast area are incomplete either because they for otherstudies and conservationpurposes: it is were basedon now partly obsoletedata (Gladkov surprising,therefore that althoughshort 1951;Kozlova 1961; Uspenski 1969) or becauseof descriptionsof thisimportant feature can be found brevity(Stepanyan 1990). Uspenski(1969) was the mainlyin regionalpublications or generalised brieflyin handbooks,there have been no detailed firstauthor to outlinethe breeding range of Dunlin in Russiawith the help of vegetationmaps, but gave analyses. no evidencefor his generalisationfrom recordsnor did he outline the methods used. A different 152 Lappo& Tomkovlch.Breeding distribution of DunlinCahdns alptna in Russia analysisof Dunlin distributionfor north-eastern densities than when other methods are used. Asia was givenby Kistchinski(1988), based on data collectedbefore the 1980s.He distinguishedthe Transectcensusing methods with fixed or variable corearea and morelocal areas with high densitiesin distances to recorded birds are in wide use in sakhalinasubspecies and paid attentionto the Russia.Data obtainedthis way shouldbe more differences in distribution of birds of centralis and representativeof populationsin widespread,not sakhalinasubspecies. Nevertheless, there is no local,areas. However, not only are suchdata detailedaccount of thebreeding distribution of influencedby the experienceand methodsof the Dunlin in Russia,and thispaper aims to fill the gap, observer, but also we do not know how data aswell assummarising more general patterns of the obtainedin thisway relatesto realfigures. Recently speciesdistribution. an attemptwas madeto checkthe latterissue (Soloviev1995; Soloviev et al. 1996). Table 1 shows Methods how widelybreeding densities may vary between observersusing similar transect methods (Ravkin Data aboutbreeding records, densities and habitats 1967) and between transectresults and those were takenfrom personalobservations from 1972- 1992 in different Siberian tundra areas,but the main obtainedby mapping.It canbe seenthat, on bulk of data was obtained from numerous VrangelIsland, Dunlin densitieswere found to be five timeshigher by Stishovet al. (1991)than those publications,mainly in Russian.Eggs and reportedby Dorogoi(1982) with both observers unfledgedchicks from the collectionsof several usingthe sametransect method. Data from the EuropeanMuseums were usedadditionally as ChukotskiPeninsula show that mappingproduces breedingrecords. The mostvaluable data were figuresseven or eighttimes higher than those foundin the ZoologicalMuseum of Moscow resultingfrom transects(Table 1). University (ZMMU), the BritishMuseum of Natural History (BMN-H),and the ZoologicalInstitute of Dunlinsshow strong site-fidelity, at leastat Yamal, RussianAcademy of Sciencesin St.Petersburg (ZIR). Taimyrand Chukoski Peninsulas (Ryabitsev 1993; Tomkovich1994; Soloviev et al. 1996;Pyabitsev & Alekseeva1998) and, as a result,there are no The precision,and sometimesquality, of the data indicationsof largeyear-to-year variations in differsconsiderably among publications. Therefore, numbersin differentlocalities: any changes are recordsof nests,unfledged chicks and fledglings,as probablydue to variationsin breedingsuccess and well asfemales containing eggs when they were differencesin departuretimes. The largest variation collected,were considered as cases of proved of densitiesrecorded was only three times (0.67 - breedingin a locality.Records of singingmales, 2.0pairs/km 2) during a tenyear study on a plot of territorialpairs, birds with distractiondisplays or 4.5 km2 in typicaltundra on theYamal Peninsula alarmingconstantly (characteristic behaviour near (Ryabitsev1993). Thus, the between-observer broods)and signsof familieswith fledgedyoung differences on Dunlin densities for the same areas constituteprobable breeding. In somereports can,in mostcases, be explainedmainly by different breedingcriteria are not stated;although in some methods of censusand calculation. Therefore, we casesexact records were used, in others this was not correcteddata for thepurpose of thispaper and the caseand sopossible breeding was assumed reducedStishov's figures by 80%. (Figures1 and 2). Thereare other sources of errorsin breedingdensity Apart from rangecontraction in the subspecies estimates. For instance the transect census method schinziidue to human activities(Malchevski & does not differentiate between local breeders and Pukinski 1983),no reliable data are availableabout migrantsin a summerbird population(Soloviev long term-changesin the breedingdistribution of 1995,pers. comm.), and usuallypopulations of Dunlin in Russia.Changes during this century approximatelythe firstand the secondhalves of a foundby comparingdistribution maps in summerseason are consideredseparately by handbooksare the resultof improvedknowledge, not in bird distribution. Hence we have combined observers.Authors of publicationsdealing with the data collected in different decades and even censusresults often do not give detailsof the status centuries. of the birdsunder study and thisleads to further uncertainty.We feel that some of theunusually high breedingdensities reported result from censusing Many differentmethods have been used for migrantsas well asbreeders. In particular,at least censusingand calculatingdensities and thismakes sometransect census data of Lobkov (1986)for comparisonsbetween studies very difficult. KamchatkaPeninsula and of Estafiev(1991) for the Densitiesobtained by territorymapping on defined north-eastEuropean tundra seem too high. studyplots should give the mostreliable data when As realdensities of breedingDunlins cannot usually carriedout accuratelyon quitelarge plots and at the be established,we decidedto simplifythe databy appropriatetime. Nevertheless,these results are assigningthem to low,moderate or high density probablyunder-estimates (Soloviev et al. 1996)and categories.This reduces errors resulting from showonly local densities unsuitable for large-scale different census methods and allows the inclusion of extrapolations.It is, likely that thereis a lower faunisticdata, where densities were evaluated only marginof errorbetween mapping data and actual 153 International Wader Studies 10: 152-169 Ira]Calidris a.