A revised estimate of the breeding population of Common Sandpipers Actitis hypoleucos in Great Britain and

T.W. DOUGALL•, P.K. HOLLAND2 & D.W. YALDEN3'

•38 LeamingtonTerrace, EH10 4JL, UK;2Rennie Court, Brettargh Drive, Lancaster LA1 5BN, UK; 35choo!of BiologicalSciences, 3.239 StopfordBuilding, Victoria University of Manchester,Oxford Road, Manchester M13 9PT, UK. [email protected]. uk * Correspondingauthor

Dougall,T.W., Holland,P.K. & Yalden,D.W. 2004. A revisedestimate of thebreeding population of Common SandpipersActitis hypoleucosin GreatBritain and Ireland. WaderStudy Group Bull. 105: 42-49.

Breedingdensities of CommonSandpipers in Great Britain (thereare no densityestimates for Ireland) are summarised,along with surveysof their nationaland local distribution,to producenew estimatesof the total nationalpopulations. Extrapolating from linear densitiesto the populationsof areasis problematical,and a rangeof populationestimates from 11,500to 51,500 canbe justified. Reasons for believingestimates of about 24,000 for GreatBritain (and about30,600 for the British Isles) are presented.

INTRODUCTION Linear densities

Sharrock(1976) consideredthat Common Sandpipersoc- It is reasonablyeasy to recordthe territoriesof this species, curredat densitiesof 10-30 pairsper occupied10 x 10 km mappingthem as pairs when they establish territories in early square,giving bounds to an estimateof the combinedBrit- May, as singlebirds occupyingterritories while their mates ish and Irish populationduring 1968-72 of 15,800-47,300 are incubatingin May/June, or as anxiousparents loudly pairs(in 1,581 hectads).Reed (1985), who presentedrevised guardingchicks in June/July.By late in the breedingcycle, estimatesfor populationsize of all wadersin Britain andIre- onemate, usually the female,has left, but theremaining one land, includeda rangeof 22,000-25,000 pairsfor Common stayswith the older chicksup to and well beyondfledging. Sandpipers,without explaining the basisof the revisedfig- Jenkins & Bell (1984) recorded birds, rather than pairs/ ures.In theNew BreedingAtlas of BreedingBirds in Britain territories,but sincetheir surveywas in this late period,it is and Ireland: 1988-1991, it was estimated that Common assumedthat their count of birds equatedto territories. Sandpipersoccurred at 15 pairsper occupiedhectad, equat- ing to breedingpopulations of 15,855pairs in GreatBritain Area surveys and 2,445 pairs in Ireland (Yalden 1993). In attemptingto revisethe estimatesfor ,for a forthcomingrevision In summarizing breeding distributions from published of Thom (1986) Birds in Scotland, we have revisited the atlases,we have ignoredthose squares in which birds were problemof estimatingpopulation sizes for thisspecies. There recordedonly as "presentin possiblenesting habitat" during are now more estimatesfor breedingdensity along stretches thebreeding season, since these are likely to havebeen birds of river and lake shore,and also more tetrad atlasesthat allow still on passage,or failed breedersalready moving back a betterestimate of how much of the countrysideis actually southwards,and this is an easyspecies for which to demon- occupiedby the species.We presentand evaluatethis more strateat least probablebreeding. Records of probableand extensivedata-base, produce a numberof rather disparate proven breeding in recording squares(hectads, tetrads or estimates,and discussthe discrepanciesbetween them. monads)are thereforecombined (Table 2).

POPULATION ESTIMATION Estimation

Samplebreeding population surveys of CommonSandpipers Two populationcounts over large surveyareas in England in the British Isles have been carried out usingtwo rather (Holland et al. 1982, Pyefinch & Golborn 2001) suggest differenttypes of recordingunit: linear surveys,along rivers averagedensities of 9-11 pairs/hectad.On thisbasis, Yalden or aroundthe perimeter of water-bodiessuch as lakes and (1993) arguedfor an overalldensity of 15 pairsper occupied reservoirs,and areasurveys based on nationalgrid squares, hectadacross the whole of Britain and Ireland,assuming that either 10 x 10 km (hectads),or 2 x 2 km (tetrads)or 1 x 1 km the densityin Scotland,which supportsmost of the popula- (monads).Therefore any attempt to estimatethe whole popu- tion, is higherthan in England,which is closerto the edge lation involves complex and problematicextrapolations, of the species'range. This gaveestimates of 15,855pairs for especiallyfrom observedlinear densitiesto areas.Moreover Great Britain and 2,445 for Ireland. studyareas are often chosenfor their high abundanceand it In relation to linear densities, Harris et al. (1995, their is vital that this is taken into accountin estimatingthe popu- Table 4) quoteaggregate lengths of rivers, streams,canals lation of much larger and probablyless favourable areas. andlake shoresin eachwater authority area. Anglian, South

Buffetin 105 December 2004 42 Dougallet al.: Revisedestimate of breedingpopulation of CommonSandpipers in Great Britainand Ireland 43

West, Southern,Thames and WessexWater Authority areas on CommonSandpiper density or presence(Jones 1983, can be discounted,as breeding Common Sandpipersare Yalden 1986), becausenarrow streams usually lack the shin- essentiallyabsent from theseregions, and canalscan simi- gle banksfavoured by this species.If streamswere to be larly be ignored. For the rest of Great Britain, there are inhabitedat the samedensity as rivers, the populationmight 28,697 km of riversand 8,591 km of lake shorespotentially be aslarge as 51,589 pairs (Table 3), but thisseems unlikely. inhabitableby CommonSandpipers. Taking a low densityof Note that streamsin the contextof this paperare defined 0.2 pairs/km,from two extensive surveys which included un- as initial tributaries,the order 1 streamsof geographers,as inhabitedas well as occupiedrivers (Cowper 1953, Vickery shownon the old 1:63,360maps of Great Britain. Smith & 1991), and 0.7 pairs/kmof shorefrom a similar reservoir Lyle (1979) suggestthat order 1 streams,at thisscale, would survey(Holland et al. 1982),suggests a population of 11,506 be 75% of the total river network,and order 1 plus order 2 pairs.An analysisby Vickery(1988) suggestedthat streams streamswould be 93%. The figures in Harris et al. (1995) narrowerthan 8 m wouldhave no CommonSandpipers, and imply that streamsare 87% of the total river network, so others have also remarked that stream width is a constraint probably correspondlargely to order 1 streams.Order 2

Table 1. Estimates of linear densities for Common Sandpipers in Great Britain (where surveys covered more than one year, the average density and number of pairs is given).

Published data

Density No. of Length River/site Year(s) Source (pairs/km) pairs of river or shore surveyed (km)

1.05 101 96 R. Lune & tributaries, Lancashire,England 1951 Cuthbertson et al. 1952 0.13 41-52 400 R. Esk catchment,, Scotland,including reservoirs 1961-1966 Cowper 1973 0.73 13 17,9 R. Noe & R. Derwent, England 1977-1980 Holland et al. 1982 0.74 14 18.8 Longdendalereservoirs, England 1977-1980 Holland et al. 1982 1.27 24 18.9 Ladybowerreservoir, Derbyshire, England 1977-1980 Holland et al. 1982 1.49 10 6.7 Derwent reservoir, England 1977-1980 Holland et al. 1982 1.97 15 7.6 Goyt valley reservoirs,England 1977-1980 Holland et al. 1982 2.39 24 10,05 R. Ashup & R. Alport, England 1977-1980 Holland et al. 1982 0.57 78 136 Peak District rivers, England 1977-1980 Holland et al. 1982 0.73 128 175 Peak District reservoir shores,England 1977-1980 Holland et al. 1982 0.45 107 223 R. Wye, Wales 1977 Round & Moss 1984 0.26 36 138 R. Severn, Wales 1978 Round & Moss 1984 0.27 18 67 R. Vrynwy, Wales 1978 Round & Moss 1984 1.54 birds a 145 birds a 94.3 R. Dee & tributaries, Aberdeenshire,Scotland 1984 Jenkins & Bell 1984 a 0.28b 39 112 18 streams,Dumfries & Galloway, Scotland 1987 Vickery 1991 (speciesonly presenton 10 streams) 1.5-2.2 7-10 4.5 NW Sutherland,Scotland 1975-1980 Nethersole-Thompsons1986 6 30+ 4.6 shoreof Loch Morlich, Inverness,Scotland 1934-1956 Nethersole-Thompsons1986 4.2-5.6 6-8 1.4 R. Dorback, Inverness, Scotland 1938-1942 Nethersole-Thompsons1986 3.0 31 10.4 part Ladybowerreservoir, Derbyshire, England 1989 Yalden 1992a 2.5 26 10.4 part Ladybowerreservoir, Derbyshire, England 1990 Yalden 1992a 1.8 60 33.3 Ladybower-Derwent-Howdenreservoir system, 1991 Yalden 1992b Derbyshire& Yorkshire,England 1.67 55 32.9 R. Lune, Lancashire, England 1976-1995 Harrison 1995 (downstreamof the site of Cuthbertsonet al. 1952) 2.47 c 21-31 10.13 streamsin Moorfoot Hills, Scotland 1993-1998 Dougall et al. 1999

Unpublished studies

0.84 62 73.6 Resurveyof R. Lune & tributaries,Lancashire, 1990 P.K. Holland England(most of areastudied by Cuthbertsonet al. (1952)) 1.36d 97.75 71.7 R. Tees,Co. Durham,England 1977-1980 Jones1983 4.67 28 6 Streamin Moorfoot Hills, Scotland(cf. Dougall et al. 1999) 1998 Mee 2001 2.41 35 14.5 Streamin Moorfoot Hills, Scotland(cf. Dougall et al. 1999) 1999 Mee 2001 0.18 27.1 153.1 31 'positiveplots' (i.e. thosewith CommonSandpipers) 1977-2002 BritishTrust for Ornithology of the WaterwaysBird Surveyin Scotland per J.H. Marchant (0.13) (includingnegative plots) 5.0 7 1.4 R. Dee at Inverey, Aberdeenshire,Scotland 2002 D.W. Yalden a AlthoughJenkins & Bell countedbirds rather than pairsor territories,it is thoughtthat the numberof birdswould have equated to the numberof pairs for the reasonexplained in the text. •' Range0.21-1.12 on 10 streams. ½Range 2.07-3.06 a Range0.30-2.31

Bulletin 105 December 2004 44 Wader Study Group Bulletin

N.W. Sutherland

L, Morlich R, Dornack

R. Dee

Esk Moorfoot Hills

Tees

Galloway

R, Lune R, Ashop, Peak District

R. Vrynwy

R. Severn

R. Wye

Fig. 1. Distributionof linearsurveys for breedingCommon Sandpipers in GreatBritain, as listedin Table 1. The variousrivers and reservoirssurveyed by Holland et al. (1982) are all in the Peak District.

streams are constituted by the junction of two order 1 In attemptingto estimatethe populationsize from the streams,while order3 streamsare formedby thejunction of numberof hectadsor tetradsoccupied, the difficulty is to two order2 streams.In the Peak District, at least,only rivers allow for the proportionof unoccupiedhabitat, by selecting of order3 andhigher are occupiedby CommonSandpipers, an appropriatedensity per unit area.The greaternumber of thoughwe haveseen them on order1 streamsin theScottish tetradatlases now published facilitates a betterunderstand- Highlands. ing of thisrelationship. Breeding sandpipers were found in

Bulletin 105 December 2004 Dougall et al.: Rewsed eshmate of breedingpopulation of Common Sandp•pers•n Great Britain and Ireland 45

11.8% (GreaterManchester) to 29.5% (SE Scotland)of the The distributionby decile of New Atlas data cannot be tetradsin occupiedhectads (Table 2). Evenoccupied tetrads deducedfrom the published map, but S. Gillings (pers. would containonly limited lengthsof riparianhabitat, and comm.) has kindly extracted them from British Trust for analysisof datafor theMoorfoot Hills studyarea in the Scot- Ornithologyfiles, enablingus to makeestimates for the Brit- tish Borders(TD, pers.obs.) suggests that an averageof 2.5 ish Isles as a whole and for each countryseparately. This pairs per occupiedtetrad would be a reasonablebasis for suggestsa total populationof 26,247 pairs with by far the extrapolation. majority (71%) in Scotland(Table 4). The New Atlas of BreedingBirds in Britain and Ireland: It should be noted that the total number of hectads in 1988-1991 (Gibbons et al. 1993) includes a colour- which Common Sandpiperswere found during the timed- contoured'abundance map' that depictsnot actual abun- visit surveyis 185 more than the numberin which breeding dancebut the proportionof tetradswithin eachhectad in was confirmed or probable during the whole New Atlas which the specieswas found during a single2-hour, or two project(1,405 hectadsas opposedto 1,220 as publishedin separate1-hour, visits. Given the time limit andthe fact that the atlas). This reflects the fact that the timed-visit survey observershad to find as many speciesas possible,not just only recordedpresence/absence and, in someof the hectads CommonSandpipers, it is possiblethat thesedata under- in which Common Sandpiperswere found during timed estimatethe actualproportion of tetradsper hectadthat con- visits, breedingwas neither proved nor even found to be tained CommonSandpipers. Nevertheless this is the most probable.Thus, although the time limit mayhave led to some extensivedataset available, and the only one available for underestimation,this seemsto have been offset by record- Ireland. ing birds that were not actuallybreeding.

Table 2. Area surveys for breeding Common Sandpipers in the British Isles.

(a) National atlases

Hectads in which breeding Hectads covered Area covered Years Source probable or confirmed

1,581 3,858 The British Isles 1968-1972 Sharrock 1976 1,231 2,816 England,Wales & Scotland 1968-1972 Sharrock 1976 1 14 Isle of Man 1968-1972 Sharrock 1976 349 1,014 Ireland 1968-1972 Sharrock 1976 1,220 3,858 The British Isles 1988-1991 Gibbons et al. 1993 1,056 2,816 England, Wales & Scotland 1988-1991 Gibbonset al. 1993 (101 Wales, 228 England,726 Scotland) 1 14 Isle of Man 1988-1991 Gibbons et al. 1993 163 1,014 Ireland 1988-1991 Gibbons et al. 1993

(b) Local atlases

Squares in which Area covered Common Sandpiper Location Years Source breeding probable pairs/density or confirmed

126 monads 18 hectads 206 pairs, 11 pairs/hectad Peak District, 1977-1980 Holland et al. 1982 England

34 tetrads 288 tetrads in 15 hectads Greater Manchester, 1979-1983 Holland et al. 1984 (11.8% of tetrads covered) England 19-34 pairsin 32.5km 2, NW Sutherland, 1975-1980 Nethersole- 0.58-1.04pairs/km 2 Scotland Thompsons1986 22 tetrads 103 tetrads in 12 hectads Cheshire, England 1978-1984 Guest et al. 1992 (21.3% of tetradscovered) 182 tetrads 1,001 tetrads in 45 hectads 342 pairs, 1.88 pairs/tetrad, Northumbria, 1988-1992 Cadwallender1995 (18.2% of tetrads covered) 7.6 pairs/hectad England

489 tetrads 1,660 tetrads in 72 hectads SE Scotland 1988-1994 Murray et al. 1998 (29.5% of tetrads covered)

102 tetrads 540 tetrads in 27 hectads - 250 pairsa, Lancashire,England 1997-2000 Pyefinch& Golborn (18.9% of tetradscovered) 2.45 pairs/occupiedtetrad, 2001 9.3 pairs/hectad 296 tetrads 1,362 tetrads in 66 hectads Cumbria, England 1997-2001 Barron 2002 (21.9% of tetradscovered) 50 tetrads 280 tetrads in 15 hectads Derbyshire,England 1995-2001 Frost et al. in prep. (17.9% of tetradscovered) a Estimate (211 counted)

Bulletin 105 December 2004 46 Wader Study Group Bulletin

N.W. Sutherland

S.E. Scotland

Northumberland

Cumbria Greater Manchester

Peak District Lancashire

Cheshire

Derbyshire

Fig. 2. Distributionof local areas surveyed(for countyatlases and others)for breedingCommon Sandpipersin Great Britain,as listed in Table 2. The Peak District,as surveyedin 1977-1980, partiallyoverlaps later surveysfor Greater Manchester,Cheshire and Derbyshire.

A converseerror might be the extent to which the New Of thesethree factors, the only onethat is relativelyeasy Atlas mapsshow Common Sandpipers as provedto breedin to evaluate is the recording of Common Sandpipers in at leastsome hectads in whichthe species was not found dur- hectadswhere breedingwas neither proved nor probable. ing the timed-visitsurveys. However, this is likely to be less Theserecords presumably relate to hectadswhere Common importantin termsof total numbersbecause, if no Common Sandpiperswere found in the lowest proportionof tetrads Sandpiperswere foundduring the timed visits,they cannot duringthe timed visits(because if they were foundin a high have been very abundant. proportion,proof of breedingwould havebeen more readily

Bulletin 105 December 2004 Dougall et al.: Revised estimate of breeding populationof Common Sandpipersin Great Britainand Ireland 47

Table 3. Lengthsof riparianhabitat in NW Britain(from Harris et al. 1995) and populationestimates (in bold) for these lengths,assuming linear densitiesof 0.2 (rivers) and 0.7 (lakes) pairs/km. The total available habitat might be estimated as either the sum of the river and lake shore (R + L) or the sum of river, lake and stream (R + L + S).

Water Authority River Lake Stream Estimated pop. (pairs) area (km) pairs (km) pairs (km) pairs R + L R + L + S

North West 1,091 218 373 261 10,907 2,181 479 2,660 Northumbria 983 197 457 320 10,771 2,154 517 2,671 Severn-Trent 1,115 223 375 262 23,048 4,610 485 5,095 Yorkshire 874 175 257 180 14,751 2,950 355 3,305 NW England 4,063 813 1,462 1,023 59,477 11,895 1,836 13,731 Scotland 20,761 4,152 6,373 4,461 117,231 23,446 8,613 32,059 Wales 3,873 774 403 282 23,708 4,742 1,056 5,798 NW Britain 28,697 5,739 8,238 5,767 200,416 40,083 11,506 51,589

Table 4. Estimationof separate populationsfor each countryfrom "abundancemap" in Gibbons et aL (1993). Figures are the number of hectadsin each decile (per S. Gillings,pers. comm.), the numberof tetrads (out of the total of 25) assumedoccupied in each hectad at each density, and the populationestimate (at 2.5 pairs per occupiedtetrad).

Decile Number of hectads No. of Number of pairs Pairs/ (proportion of tetrads hectad tetrads England Wales Scotland Ireland Total occupied England Wales Scotland Ireland Total occupied in out of 25 hectad)

>0.624 1 1 139 14 152 18 45 45 6,120 630 6,840 45 0.500-0.624 1 0 117 14 132 14 35 0 4,095 490 4,620 35 0.375-0.499 15 5 115 23 158 11 413 138 3,163 633 4,347 27.5 0.280-0.374 36 4 78 4 122 8 720 80 1,560 80 2,440 20 0.250-0.279 7 7 100 35 149 6.5 114 114 1,625 569 2,422 16.25 0.167-0.249 47 9 60 10 126 5.5 646 124 825 138 1,733 13.75 0.125-0.166 52 26 90 92 260 4 520 260 900 920 2,600 10 0.111-0.124 20 9 13 15 57 3 150 68 98 113 429 7.5 0.077-0.110 44 17 25 13 99 2 220 85 125 65 495 5 0.040-0.076 84 16 15 13 128 1 210 40 38 33 321 2.5 <0.040 16 6 0 0 22 0 0 0 0 0 0 0

TOTAL 323 100 749 233 1,405 3,073 954 18,549 3,671 26,247 18.25

obtained). On this basis, if the 185 hectads with the fewest tetradscontaining Common Sandpipers were discounted, the Table 5. Populationestimates for the BritishIsles, based on likely estimate for the British Isles would be reduced by 1,087 proportion of tetrads occupied (at 2.5 pairs/tetrad) in hectads in pairs, from 26,247 to 25,160. This discount, however, is which breeding was confirmed or probable according to the New likely to be morethan offset by underestimationarising from Atlas of BreedingBirds in Britainand Ireland: 1988-1991 (Gibbons the time limit. Thereforein our opinionthe bestestimate of et al. 1993). the populationthat is possiblewith thesedata is that set out Country Hectads % tetrads N tetrads Population in Table 4, but it is possiblystill on the low side. occupied occupied occupied estimate A similar approach,also giving separateestimates for eachcountry, is offeredby counting(approximately, because England 228 15 855 2,138 hectads,tetrads and monads straddle national boundaries) the Wales 101 15 379 948 numberof occupiedhectads in eachcountry, combining the Scotland 726 38 6,897 17,243 resultsof local tetradsurveys and the occupancyrates from Great Britain 1,056 8,131 20,329 Table 4 to estimatethe likely occupancyin eachhectad, and Ireland 163 23 937 2,342 applyingthe sameassumed density of 2.5 pairs/tetrad.In Man 1 10 3 8 Englandand Wales, it is likely that about 15% of the tetrads British Isles 1,220 9,071 22,679 are occupiedin eachoccupied hectad (observed range in six English studies,11.8-21.9%, mean of 18.3%, see Table 2; mean rate nationally from data in Table 4, 15%). There is only onelocal Scottishatlas in Table 2, suggestinga higher occupancyrate of 29.5%, butthe analysisof Table 4 suggests

Bulletin 105 December 2004 48 Wader Study Group Bulletin on average38% of Scottishtetrads were occupiedand 23% assumptionsof 2.5 pairsper occupiedtetrad or 1.6 pairsper of those in Irish hectads.This gives an overall population occupied monad would come from more atlas surveys in estimateof 23,000 (Table 5); this is slightly lower than the which pairs/territoriesare counted;the latter mustbe at least estimatefrom Table 4 in part becausefewer hectadsare rec- one pair, giving a minimum populationestimate for Great ognisedas havingbreeding pairs. Britain of 15,300 pairs if the occupancyrates are correctly The best stratified sample of breedingCommon Sand- estimatedby the BreedingBird Survey. The populationof pipersin theBritish countryside is theBTO's BreedingBirds the British Isles would then be at least 19,500 pairs.Because Survey(BBS). Mead (2001) reportedthat this foundCom- it is basedon presencein this stratifiedsample of the coun- mon Sandpipersin only 3.4% of the monadssurveyed over- tryside,the best supportedfigure seemsto be the estimate, all in Great Britain, but in ScotlandM. Raven (pers.comm.) usinga densityof 1.6 pairs/monad,of 24,000 pairsfor Great reportsthat the simpleaverage over 9 yearswas 12.9%. To Britain (including19,000 pairsfor Scotland),30,600 for the clarify theseproportions, S. Gillings and S. Newson (pers. whole of the British Isles (Table 6), and it is reassuringthat comm.) have reviewedthe BTO data for 9 yearsof the BBS. the estimatesin Table4 - whichwe expectedto be on thelow They calculatedweighted averages for eachcountry, allow- side - are similar. The exerciseemphasises the importance ing for the differentdensity of coveragein differentregions. of the Scottishpopulation, not previouslymade explicit. It Based on 1,558 squaresin England and 174 squaresin also reflectsthe availability of freshwaterin Great Britain Wales, only 2% were occupiedby Common Sandpipers, describedby Smith & Lyle (1979) who show that 66.7% of contrastingwith 15% from 273 squaresin Scotland. No the lake area and 52.9% of the rivers and streams are in Scot- squaresin NorthernIreland or theIsle of Man wereoccupied, land whichsupports 79.7% of the CommonSandpiper popu- andthere are no equivalentdata for Eire, for whicha propor- lation (from Table 6). There are no BBS data for Ireland, so tion of 5% was rather arbitrarily assumed(scaled up from the Irish estimatemust be takencautiously. Moreover that for Englandand Wales, down from Scotland,cf. Table4). There the Isle of Man is clearlyexcessive, since it is rarelyrecorded seemsto be only one publishedsurvey of an extensivearea breedingthere. The estimateof 30,600 implies a densityof at the monad scale:Holland et al. (1982) found 206 pairs in 25.1 pairsper hectad,which might be on the high side(Table 126 monads,at a mean of 1.63 (s.d. 1.09, range 1-8, median 4), and is substantiallymore than the 15 pairs per hectad andmode both 1). Our two studyareas, selected for theirhigh assumedby Yalden (1993). populations,contained means of 2.56 (T.D., pers.obs., Scot- In sofar asthis analysis is basedon the 1988-1991 hectad tish Borders) and 2.35 pairs per monad (DWY, pers. obs., distribution,the populationestimated is essentiallythat of Ladybower/Derwent/HowdenReservoirs) in 2003. It seems about 15 years ago. During 1968-72, the populationmust likely that the averageper occupiedmonad is nearer2 than havebeen about 39,600 pairs in the BritishIsles (30,900 pairs 1 pair (since totally unsuitable habitat has, at this scale, in Great Britain, 8,700 in Ireland), assumingthe sameden- already been excluded), and using 1.6 per monad gives sity acrossthe wider distribution.The populationapparently estimatesof 24,000 pairsfor Great Britain and 30,600 pairs declinedby 23% betweenthe two nationalatlas surveys, and for the British Isles (Table 6). continuesto decline.Analysis of the BTO's WaterwaysBird Survey also suggestsa decline of 23%, between 1989 and DISCUSSION 1999 (Baillie et al. 2002).

This exercisehas generateda large range of population ACKNOWLEDGEMENTS estimates, of variable reliability. At least, however, the assumptionson which they are basedare clarified. We thank Drs M. Raven, J. Marchant, S. Newsonand espe- Underlyingthe variabilityof the estimates,there are two cially S. Gillingsfor accessto unpublishedBTO data,and to centralproblems: Dr Gillingsfor valuablecomments on a previousdraft of this paper.Drs S. Jones(Shirley Goodyer), A. Mee andJ. Vickery 1. What proportionof habitatis suitable? haveshared unpublished data and details of studysites with 2. At what densityis it occupied? us. We are very gratefulto Dr H. Sittersfor thoughtfuland It is crucialto assesscorrectly the numberof pairsper occu- constructiveediting. pied hectad, tetrad or monad. More confidence in the

Table 6. Estimates of the populationsbased on the proportionof 1-km squares (monads) reportedoccupied by the BritishTrust for Ornithology'sBreeding Bird Survey in England, Wales and Scotland,applied to the areas of each countryat presumeddensities of 1.6 or 2 pairs/monad.

Country Area % monads N monads Population estimates (km2) occupied occupied at 1.6 prs/md at 2prs/md

England 130,298 2 2,606 4,170 5,212 Wales 20,824 2 416 666 832 Scotland 79,245 15 11,887 19,019 23,774

Great Britain 230,367 7 15,361 23,855 2,818 Ireland 84,364 5 4,218 6,749 8,436 Man 563 2 11 18 36

British Isles 315,294 6 19,590 30,622 38,290

Bulletin 105 December 2004 Dougall et al.: Revisedestimate of breedingpopulation of CommonSandpipers in Great Britainand Ireland 49

REFERENCES Ecology, Banchory. Jones,S.A. 1983. Ecologicalstudies of wading birds ( Charadrii) in some Baillie, S.R., Crick, I-I.Q.P. et al. 2002. Breeding birds in the Wider uplandareas of Britain. Unpubl. Ph.D. thesis,University of Durham. Countryside:their conservationstatus 2001. BTO ResearchReport Mead, C. 2000. The state of the nation's birds. Whittet Books, Stow- market. 278. B.T.O., Thetford. (see www.bto.org/birdtrends/). Barron, P. 2002. CommonSandpiper Actiris hypoleucos, pp. 152-153 in Mee, A. 2001. Reproductivestrategies in the commonsandpiper Actitis Stott, M., Callion, J., Kinley, I., Raven, C. & Roberts,J. The Breed- hypoleucos.Unpubl. Ph.D. thesis,University of Sheffield. ing Birds of Cumbria:a TetradAtlas 1997-2001. CumbriaBird Club. Murray, R.D., I-Ioiling, M., Dott, I-I.E.M. & Vandome, P. 1998. The Cadwallender, T.A. 1995. Common SandpiperActiris hypoleucos,pp. BreedingBirds q[ South-eastScotland: a Tetrad Atlas 1988-1994. 94-95 in Day, J.C., Hodgson,M.S. & Rossiter,N. (eds). The atlas of ScottishOrnithologists' Club. breedingbirds of Northumbria.Northumberland and TynesideBird Nethersole-Thompson, D. & M. 1986. Waders; their breeding, haunts' Club, Newcastle. and watchers'.T. & A.D. Poyser, Calton. Cowper, C.N.L. 1973. Breedingdistribution of Grey Wagtails, Dippers Pyefinch, R. & Golborn, P. 2001. Atlas of the BreedingBirds of Lanca- and Common Sandpiperson the Midlothian Esk. ScottishBirds 7: shire and North Merseyside,1997-2000. Hobby Publications,Liver- 302-306. pool. Cuthbertson, E.I., Foggit, G.T. & Bell, M.A. 1952. A censusof com- Reed, T. 1985. Estimatesof British breedingwader populations.Wader mon sandpipersin the Sedbergharea. Brit. Birds 45: 171. StudyGroup Bull. 45:11-12. Dougall, T.W., Mee, A. & Yalden, D.W. 1999. Recentfluctuations in a Shatrock, J.T.R. 1976. The atlas of BreedingBirds in Britain and Ire- CommonSandpiper breeding population. Scottish Birds 20: 44-45. land. British Trust for Ornithology,Tring. Frost, R. et al. (in prep.).Breeding atlas of theBirds of Derbyshire.Der- Smith, I. & Lyle, A. 1979. Distribution of Freshwatersin Great Britian. byshireOrnithological Society, Derby. I.T.E, Edinburgh. Gibbons, D.W., Reid, J.B. & Chapman, R.A. 1993. The New Atlas of Thom, V.M. 1986. Birds in Scotland.T. & A.D. Poyser, Calton. BreedingBirds in Britain and Ireland: 1988-1991. T. & A.D. Poyser, Vickery, J.A. 1988. The e•kcts of surface water acidt,'ficationon ripar- London. ian birds - with particular referenceto the Dipper. Unpubl. D.Phil. Guest, J.P., Elphick, D., Hunter, J.S.A. & Norman, D. 1992. TheBreed- thesis, University of Oxford. ing Bird Atlas of Cheshireand Wirral. Cheshireand Wirral Ornitho- Vickery, J. 1991. Breeding density of Dippers Cinclus cinclus, Grey logical Society. Wagtails Motacilla cinerea and Common SandpipersActitis hypo- Harris, S., Morris, P., Wray, S. & Yalden, D. 1995.A Reviewof British leucosin relationto the acidity of streamsin south-westScotland. Ibis 132: 178-185. Mammals: population estimatesand conservationstatus of British mammalsother than cetaceans.JNCC, Peterborough. Yalden, D.W. 1986. Habitat and activity of Common SandpipersActitis Harrison, K. 1995.An Atlas of BreedingBirds in Lancasterand District. hypoleucosbreeding by uplandrivers. Bird Study 33: 214-222. Lancasterand District Bird Watching Society,Lancaster. Yalden, D.W. 1992a. The influence of recreationaldisturbance on Com- Holland, P.K., Robson, J.E. & Yalden, D.W. 1982. The status and mon SandpipersActitis hypoleucosbreeding by an upland reservoir. distributionof the CommonSandpiper (Actitis hypoleucos) in the Peak Biological Conservation61: 41-50. District. Naturalist 107: 77-86. Yalden, D.W. 1992b. The Common Sandpiperpopulation of the Lady- Holland, P., Spence,I. & Sutton, T. 1984. BreedingBirds in Greater bower reservoircomplex - a re-evaluation.Naturalist 117: 63-68. Manchester.Manchester Ornithological Society, Manchester. Yalden, D.W. 1993. CommonSandpiperActitis hypoleucos, pp. 192-193 Jenkins, D. & Bell, M.V. 1984. Vertebrates,except salmon and trout, in Gibbons, D.W., Reid, J.B. & Chapman, R.A. The New Atlas of associatedwith the River Dee, pp. 83-93 in Jenkins,D. (ed.) The BreedingBirds in Britain and Ireland: 1988-1991. T. & A.D. Poyser, London. biology and managementof the River Dee. Institute of Terrestrial

Bulletin 105 December 2004