PN a a ae ee

ae

AAR PAE Ae WWACRO re MVA: Me AP

BIR AA MESS

tS ed De Be

Ped A ae Ra dhn #

ied bouke tan Sul Et yen ven ie Wao Te Co ee a

Mh ebes A ome Deveewe

AH me RPE BTR sem sagt att Bnet ner MOF ERO ENS kt BMH Ee ASIN I DE LPs ote hee &% wens SOLOS tees tae 2 ee Re ee ee eet’. Pettit) Slo et re ee Tere tr ro on. orn Tey MR OE BN Br 8 ge ge FM AME MAME Ra BPM Be ANTONE Nem gite eh MEET Bie Pree h id OM atop Mfr ohh eaten oe rere ow Atel 2 PORN DT ea teem eatin es Daas Pera ee ed MOE IIE NCE a te DM St Mtn tan A fe man) Ae Ve tA Ae Ae ar duennins a <A PAE eM Bede et a AmB ATRL NOE Petter ck cat LT ADL ANA Mh LB A Rte b efi ho lh DEN a Ld Lf BS ar ded in flit ee kis ee em a ot a Pee ed ee es uerey . NE te NN RE MANY be ter at ah BARR BINA DIE OS Mon KP ore N PMT MMe) Bin B DL elodl have DPE NM DALE ODA per Das Aten Be ID Feet De ol thon AN Oe Obs ee te 2 ON TONS ., Sr OR Rees ke cet ee ti Nt tea ty wt ee ett nAAM o At n Katte SAAB Le CRAB ARE RAR BOE OLA IIE NET PSEA Be Bi foe HEU bien % . ‘ Tah tie, Bilas tal Sys biota C pe okt Ranta ana Pee Pe eee ey ON TNE ee it OUT ATO en tts ow tm em 1S BA eR netlist Sh Bd AUN NL AD ALR EERIE hi GMT ERE HD Al ahh EF OME ALI E M ML ke PR AM AOR LB hE nol pet? OP es ton Seti ee name te Aa Ane: To eek netted geile died saved CAMA AR SEMEN IRI Mb hh Kody OE emt Rohont tot hatel OND doth MA tale WE Vet r ‘ nie ott aes er eee SUR ATR Artem eke SR PPT Ae eee ALPE A oan Pap ab AANA ally IOS bw Teonbootalfirobee lho” Sia Tet fete ty Al elif Ih Lele ink Shoot wet : saat tony ri. Weer SOMERS Nt ETN TN ETRE Re em gy NW RT te Men ems Rony he ae Si CYS al ey eet NPY Ce he ee ee ee a oats + Key See ashy ensahahader mee aunt Or) a apy I Re en OL ee Po itiat tek WAL fen ee tinh A CET tet oe oe a eeeren ye Pe en en een ee eerie Pron er ee Cray y Hav stars md ofboth ak saree Kita Min thn t HDMI te dA Srermtierfn fh En ARMA oie aL shine at caren) a . tren y SR TRENT tT im tie Na Retina le Moet Me RG my OREN Tem veda aitinZsy Pine sA’rtnighe of'ek ate OO STUMPED OTE An Bs planters RntONE tte Rotten tinned etal: 6 I 4 ‘ ‘ * arntabn Ta terhARt he Oa). ladle Made ds nang haat AMARA anes *95 SSR LSTA ak och Ane a CON AO ge Male OME fe Mor Sant ment on Se ree wer et en Te shes Ven Ake ANd ok Nn IAe RRs OM MANING tee doen » eRe Me Ane ee AS a eee ne Oe y / Lraclited® 2 a r Peed, Bes ATR ADA ATR ER Met MeN ee ety s Secrneny erecer rns Crepe ree ae ee ee me ee ee oe eee yet bbe Seren Age Autiiei Ades" het Lathetaptae cease Pererereryr cre ee Be . 4: SA Sone N oma nt ee et ine NEI AE AT PEEP BL Ot LATE 9 tae diab met Moms ren eae soasagages ‘ a SMS T AA rear irs CNN eh ON Bd MRS Donte nt us nereny opbiaieibiles ‘ sone ee eerie See te Seren Teen or Pr tt i rere a atenuthes, le PON re oe ER EO Le Ahn PM ial hap TE nee EN REET At eR ERED Land shLinestinad iio NetguSedeerutacanishitedis (angi Plkais.- sept man mmr nsinetaieagmaaieetaliie a ates STN TEE MOR RR Neen eget nett Nome nemo mat tmnt en wr Oak A EAN A DARIO FEB AF Dnt de Fa git gM As ne Bata hoet tae i fhehiteg oF en ee Lote totot eke me See 6th Sm Serer) De DN Tey ee rene en Peer ea re enter eee ee hr DOR AN wel OF og oe at APO 8 ter | he Sea ee Pheer ’ ’ ‘ ’ Dak son Th hee '% mo Mate Made e RTM RTD Oe Ae Me een Mae, Bae Bere teh heme ee

aes

PALL ane A eR PAE L AMA SE pet Ot ot Mn LW Rinky PN ENS WF biee Lignin gat Ped to > Oren erent wy een some eAAARSmAn ts Maw a ta Scam sua tame ees cies eM ate", Aaa AAR at AC SIIEE ee led pr eBh bel wh) Wak BRA Binds Pia wt tims vethin rn maceteo cat ThE Dot ao ‘ tama ERY AE . athe sus bes © ARN ete Pe LaRaA Leg ancDL.04 LNA RAAB SERA) B9K Seer err PA att nastier gia Aegean ni IFS ele PymAo sin 7 a Po TO ts blges "2 Nepasbonbiya Mut Hojpmt rare? GEN oantenes toner Cr rete | ee ror ee ns Nene eee APM e Rad tarts De ON De Rak eon rma te Sy een ee Optee tor WO EO at a tied PETE POE het ot iF abs bree BN Fiano, Nite tines spstanaiien Lh uermey Aeita Se Re, ren, Vale aca nea naardette hatsmole ta Me asa, Ease ae nk Menta

oN ema aE te MINA hl A sheotitian tn Pintle tind nt Siva ee afin Naty! atin assay Mycet ! SE tah ia? le rs en Doane <a MBAR Anasac Aa miases semen athe EMAL AH SIAN es 5 7 a ar) ’ Coe eee Y . . ele. ~ woe Tee ee te aE eer ee Ty a Oe Th OT Ue eae eT " ne rene ay ‘ ete dette Rants, rr a8 ‘, aaate eT er ee ee ee ee ee ee Pat: ntiaan 0S Pint 14 eh sMeSebse toi , 3 ou al a acti PliDE® ih Mea Moar o, ere vere een eres PA ue eT pol GA mie ee BOR DAE A mt FM EIB Pe aria tl DE pene ie tial tim Oe Fer Pn dt tt , ‘ ' Piano: ne Pont Sate ast Noh 0 Da gnnane toannany se ae Roh Me LES PAUL Haw eee ee ee e ve ee er ae - ‘ pom dee Nth pear Pre ere terry Cg Sin alae tne meh te A a Ft ako We ‘ . Cron ere re ENTS Oe PN ae na ABA, A O43 8E Ot tee ALD ate Fe othe oe aie gal sg LiDisind i Min pe Ny te TA STN ee Ree nate ae Ste eRe Reem ® POA ee pga AE Nk tt eto he ~ vor , Teeth arta . acasPane ye Conan tats tons , ’ ‘ afi Bo yi : Sotasaatkedente Vovenwdenes a ee, ee ee re as an

re ee ee es

SO Ne et oe Cot

mete

be Che Aaeip time ntialh 4) 6 falls Ome biltet Yasmin tape vt os " ‘. sis oe Lees AMRORNS Se he Ree ed es ee ae ‘ ia hey Ste eed caf aimnd 09 ieee eon Hf ate? . 9 weds ete lnarty de N55 a mete Sek men > ote hel 0 a ntiidime AO at tet ad wheal ots ry ' . Vette he 2

Let te toe

ee RH tat ome

ON ae ee ee ee

Beer Tron it rer See fand atban Neal Cattle Cate a tiearst ia re pe: tke A me nt eee carr 7 Lihat PAA CEMA TS eh NEON eA er ws a een eet ere Wey ee ek ee aL OO oath a her ry Satie nite hee OL OtA eh Ae He oa all Sot Mb = whe - a) ‘ akS One See R ah Se Marte Neher te tele iad er Peery ee te ee ee eT ‘ ot oe Nod ful al yt Line Glsi teitiins 3 icdeessarys . sdb nicl Matias, Mash pathos ail aslie Rigel OT er te Pe el Jibs ise fn ch ssh Se Nanbidenl cod t ‘ ie rn er eerery rn eee rene rer ry Poe oT. FokiaD Asaeinptialinastms plentinpl taba tides ina saci seu sabes Vesta 2a FNBNY ge 4 toe ae ready oiling Apt ae a semen in Naas ina esaeit emia tiaadieaiee - Sener ene ' See ee set Fe Shea an ot nent. “at Neo hd STON betes en New Oana tebiintniah ia? oA RAO La! Asad oot nt ith Melitaldettinee a) . : fr ne A ' VA Acinw Seka cane Pn PET SRE ES IS ft Cm ef PCr ved : 7 ty 2 ION Aerts er? roe " top 8 AL SRO E ppl dane ,d if estates TT dla Cr ee ee Pernutie oh Te 0 cntimteB ‘ bane: js pri zFeriy BW fae Neti ots Pine Ptalie Diath sliyelety 6:50 340 Si lone fh lala? ges Sh. aes fata Bec mrs rE REL OY Yas CE TS REY , neers SRT aya tier rar nebulae aera .

pint VA atdel ech at athe Con ee U e dat saat ‘ , ne 1 a vy, “Af ss? habacanas: (ALORS A Sepa oop Pall ay Saad indindknk " ror Airey 3 ata pore Abadia fe o.in 2s £Ah7i eM Ebe rm Dre Sear Seen er etc ae ified wef Se , ah featniends Sire Ene Ae esp eres: * : cea eel be ¢. a ‘ , Of nt . Nat, . io) ween € ‘ Ms raliie tier ra A ean . fas syste ’ ‘ ’ Dive eh we? nat ota ak Ili petogea ml an D eg? 4A ti Peddind oth dhe Cee 74 Ft fiottog og ; Lash ‘ ‘ a isdemnaiemnasanm ta ' ‘ ee i Ramee pe Bar mransdmialien at RP aC inet tee ate et ¢ toe poe eae ¥ 7 : t cre ‘ ‘ Yaeeeu? iSialingl ‘ Pagan tne ~ Pers ipo iy Retest: oad Py tnt =, pepe i tern —a te ‘ nee ee ee: ma hea FA Dent an oF ak nEies! oF ACEO? po Fee Poe eet oe "s 4 Ph ahi Sogmo teh . Feathot 4 , £ are oid , ' el eae ears a = " Yt hithag ‘nigel Maree ES * stortoeg Bat ey ir FO ene eee . . Pam Mane Ain le ss dcannallis wah uti atetiaalte.” nde * Petal Ie Prion ’ Oe Ramee SP af fee wo . Peutentind) : ; 7 A 7 aie iret, Aetie ‘ Ki. Be " shia A —— 4 + Cie a A fhe nenae : =. teases mame on Atenas eyer RP Ra ERR topes “ 6 r he i nai 5 veils aah ae ar 7 thy Ce apr > x4 aun ca ent eter See ever ~ oe sete nenete oe Fy eglerir sm dtoes. Cpe ety mien 9 oP . tides! Hahn ret. Lindi adeas Fuge oi ihe cauiaes a iy 3 aires Hie sae fone yes ae ha"aptin Selithatiotnes tant’. te 0 haithen Real enfin Son Rep” - SE en 2 meet Fi 8 hae nah Pe . a) tees roe Phin Pagenal ate iad iat p os . nea ’ 3 ; rs remit pa ‘a “i a ‘ . <a5 alee aes anes m 7 a tet nete . ‘ yu - " nas Piling ® ta ‘ ' . ' “nal fa . Sait, rye a ee ae tee ee at eer mg Seen Ma Niain ‘ ~ teeta ete te MS A ~otetha mon © Bee cat heai las» ab ind at plhaal write woes eure é ‘ O64 aoa Lent cit Pus Dasmaist ’ eae hy i Poe : shed nae 7 LOAN etal . man i re be oa pi a a"meee rere attends i? tad th a ‘ vo Font at al Pattee ly . ate Reais i Pepe : . x Es Bi, ert, wr - . 3 7 wes eee prea . Fas "i i aaa ferear at + Pwr 0, ton ene - j Un et Taegan t ; Fi nerdy . F ae ets s 4 Ms : Loe ee Sms Tee te et Ragen 96 eee N P a cae , cots . oe . Raglan Pas ' , ; ; . Faieede . - 2, 95 stds ali ae sae ae 7 - , eee ke cs : rte tani ‘ nA edie Sete 'N ma . , E pes e nt i itm Pre he . ‘i ners ie : : . re aPine posh . . ae ne ast Milan ai fauna , rs a fn hs . ‘ c _ Rane = . Gre = . - . hase ° - . ‘ Metadata Tsp Sica pills casl . Ve ce eee pial : . . artes “ . Cait ’ ' ’ . A ; Mehatons ‘ Pig hs ‘ >, ae - Paonia if i aie akan eintine ih ‘ . : one AW art an ” : meat Pian ted tt : $y ven) Pn . . ‘ , SMe Nat ae as nl 7 ee 5 pir s ahupien's harm a a men te “* mente St rie rUms dian An ag? iv nthe? 'AP Anafitse n* nthe . one : ee "0 ‘ toe + en Pe rate , r - : ‘ ae . 7 site 5 Fag . . “ . eArene > ‘ one - 2 * eee ns st wee Pe tht . ’ . oat at bo ; ' F tas, ‘ A cua . ‘ i : . . ee = natant steos Ste “ jlo ‘ Rar 1 yr er oO . A . » ‘ " ar J Rai Aerts 2 4 4 ? “ men . -* . = sina pains a Ry (etn ne f oo F ‘ Me tad . eB k fe bh 4 bn A » . a Setar pa ' ‘- . . ” . ’ , , 7 y nit ical m P nae ene ” nn easfare ein ms . . yr P " % om . ® ’ . wee cane ' ’ * . faa , ‘ , e : 3 a wn wont Ayrtiny Meets Bota potas Ye tale lgeinty ore ae ee es ponds owned ee ee ee viet en F ¢ = 7 y io men ater : soe coe SHR ped os Pangee , = epee nthe Cet sate _ Sle ag te EN at ee Quinney one de r Dabs : A wpe eroevenmee = ble ay aye Ne De EES SOs ee hd hs a a aw © AL, Ab AL MP bd Led ae oie Dot ae Loe PRATER CoO A PRA Oa Moar ne St

prt ae way >.> a ey

SMITHSONIAN INSTITUTION NOILALILSNI

NVINOSHLINS S3Iuyvudi7 LIB

RAR’!

5 = = a z é ‘oO = w = wo 2) aa ae > rad ae ro > er ee A ras 0] —_ a f) eh W”) _ Fe oa > rr z m m = ” = W” a z ” SAIYvVUdIT LIBRARIES. SMITHSONIAN NOILALI 2 2 ae ae z z ; = je AS = = 3 s > ear '@) INS a po 6 oO x= Vee, A on DM WY™* @ D O vf Y Ee O *% ms We 2) a Oo pp = = os Y = = = WG > a = ee coe = > “G A a 7) eer wn % Te SMITHSONIAN INSTITUTION NOILNJILSNI NVINOSHLINS S31uvugi7_LIBRAR t z wo z fy Zz ue 2” coo wo ae ” iu = © a oc ps x % = <x a <x = : = a x fo = : Z o = o 2 3 S| a | ae a Pe = rs NVINOSHLINS S3IYVYGIT LIBRARIES INSTITUTION _ NOILNLILSHP a cS a = Ea é | - & 2 = 50 z = = Mae ; = = rE a ss : E ze i= e ES ee D z D = ,SMITHSONIAN INSTITUTION NOILOLILSNI_ NVINOSHLINS S3luvud!7_LIBRARI Ze z o, < Ste ae 4 z 4 Vt py, = 4 z . : 2 fe * g = A > = a = >" Se a ” = 79) z 77) ni ; LIBRARIES \SMITHSONIAN INSTITUTION NOILNL 7 ”) = ” = 7) n = ow MY) = us ; oc = oc pes ox : Sues = < ae < a = a jae 4 m& 5 = re) ra rat fe z | ~ = _J =a ad SMITHSONIAN INSTITUTION NOILALILSNI sauvadi7 LIBRARI a - z a Zz = 2 - OO = ow = ow / = = 0) > 0] a) > = > | > > : a Ne, a = Po] ra ts ; y, a ” as “” oc} a ae a z Oo z oD _NVIMOSHIINS (Sa IyvuUg Ae te INSTITUTION NOILALIL < SiN aa = 2) at ae | z = 4ge7 5 z Z Wo& > aan a 2 se NOE ONE : 2 “ig = Z c > 2a ; =f eae z - § 2 SMITHSONIAN _INSTITUTION NOILOLILSNI_NVINOSHLINS Saluvugi7_LIBRARI

NIAN INSTITUTION NOMLIILIZONE NVINUSHIING salva gil] Fi PC RANITES

1G = = = - é = o e 2 > = > ‘ = > = = = Y o Z 4 ee a zZ OSHLINS S3IYVUAIT_LIBRARIES, SMITHSONIAN INSTITUTION, NOILOLILSNI

: = ix aS = A = = aes = a = age | 2 RR Ag 2 As w) w wo O :- aie oO oO zZ = z = z a = =

INSTITUTION NOILNLILSNI NVINOSHLINS Sa3iuvugi7_LIBRARIES

= < = oO aS a ore = 22) —_ wo = a ae uw z nag a z wn” Ass Ww fet Z “4 NS = = te < 4% Was < pa. 2 2 3S MQ = = x S — g JOSHLINS S3IYVYSIT LIBRARIES SMITHSONIAN_INSTITUTION = = 2 Z ee = SS = a = dy aa = \ b zi E Pe Za eee = ms ie _ op) = Ww =_ THSONIAN

INSTITUTION NOILNLILSNI NVINOSHLINS S31uVvuall

NWINOSHIING CAIIMYHAIT LIBRARIES SMITHSONIAN

= z = z 2 , < 2 Ng : g : g 2 a 2 = Z 2 SS E Wr 2. z 2 = NO ie ; = a pod YVNGIT_LIBRARIES SMITHSONIAN INSTITUTION ,, NOLLOLIISNI_ Ps “ = Yin, oe = 5 ae A < Gy ft A < ‘es SN a ro) m. 5 3 c ; i. i a ee 3 > a) < MITHSONIAN INSTITUTION NOILALILSNI_NVINOSHLIWS Ssiuveaii LIBRARIES | * =. ot, rf F z we a be ~ o = 0 : r 2) = m - 20 - = : 2 = = ; = - = = e . S 2 m i : : So = wn = SHLUINS S3INVYGIT_ LIBRARIES SMITHSONIAN INSTITUTION NOILALILSNI _- ” rad oe wo” ua oo ~” < . = Ae RN = ae "i - S OS = 5 Ree ¢ n YM NWS : w Tt) N (op) U & se ® oO in oO J = = YQ = ae = . eG \ 2 a ee ee _INSTITUTION NOILNLILSNI_NVINOSHLINS Saluvy giq_LIBRARIES = = > So > SOM :

ne } i} f \ J { | in Ve \ F Find ne i} ia! Pa. i Pare De ty i hs ‘ ie Hy iy uo im +

t . 4h At 1

ih

OO hee,

Ti

4 5

hans b i alll ly (eA = oe iw

SS 4 \ GY I! sy GY | San" L

ha = “gf y)! Di.

oe

Scottish Birds The Journal of the Scottish Ornithologists’ Club

Editor: N. Picozzi Assisted by: N.P. Ashmole, M.A. Ogilvie, S.R.D. da Prato and Valerie M. Thom.

Scottish Birds, the official journal of the Scottish Ornithologists’ Club, publishes original material relating to ornithology in Scotland; papers concerned with status and distribution are particularly invited. All papers are considered by an Editorial Panel and, where appropriate, are scrutinised by specialist referees. Authors are advised to invite comment from friends or colleagues, and if necessary to make amendments, before submitting their papers. Short notes on unusual observations or records are also accepted. (Advice on the submission of contributions will be found on the inside back cover.) Papers and short notes should be sent, in the first instance, to The Editor, Scottish Birds, 21 Regent Terrace, Edinburgh EH7 SBT.

Two numbers of Scottish Birds are published each year, in early June and December. The winter number contains the Scottish Bird Report, which includes rarity descriptions (these should be sent to the SBR Editor: Angus Hogg, Kirklea, Crosshill, Maybole, Ayrshire KA19 7RJ).

Scottish Birds is issued free to members of the Scottish Ornithologists’ Club , who also receive a quarterly newsletter, Scottish Birds News, which carries items of topical interest, Club news and notices, and book reviews. Details of SOC membership rates are given below. Scottish Birds and Scottish Bird News are available to non-members at a subscription rate (1986) of £25.00. Binding arrangements are available for Scottish Birds; volumes cover two calendar years. For information on advertising contact the Business Editor, at 21 Regent Terrace, Edinburgh EH7 SBT.

The Scottish Ornithologists’ Club was formed in 1936 to encourage all aspects of ornithology in Scotland. It has local branches in Aberdeen, Ayr, the Borders, Dumfries, Dundee, Edin- burgh, Glasgow, Inverness, New Galloway, St Andrews, Stirling, Stranraer and Thurso, each with its own programme of field meetings and winter lectures. The Club’s headquarters are at 21 Regent Terrace, Edinburgh EH7 5BT. The Waterston Library (probably the best ornithological library in Scotland), is available for reference during office hours (Monday to Friday, 9am to 5pm). The SOC Bird Bookshop is one of the world’s leading ornithological booksellers; it carries a large stock of English language titles and can supply any Natural History book in print. Much of the business is by mail order, but personal callers are welcomed during office hours. The current catalogue is available on request. The Bird Bookshop is run by the SOC for the benefit of ornithology in Scotland.

SOC annual membership subscription rates

GUNG 5 8 oleciat Has eee bersoeineteale aavocialtrmnie telecine stvenioa st nnemam eens. aan een Ae Oe ee £10.00 Junior (under 18, or student under 25) .......6.0...6¢.ccccesseseee0e00 ce oes Cech eee £4.00 Family (husband & wife, S& children under 18) .....55....45:2--.+220 02. £14.50 JEWS JGACIMUCDIO! scoocpccscccoaa09o0c006e £200.00 FQMULY oocbsenenacosesecne eee £300.00

Pensioner Individual ................... £6.00 PQUNUY so acted ccuinoninn viene eee £8.50

Scottish Birds (1988) 15: 1-8

The recolonisation of the Isle of May by Common and

Arctic Terns

S. WANLESS

After an absence of 21 years Common and Arctic Terns recolonised the Isle of May in 1979 and 1984 respectively. The colony increased each year and in 1987, 76 Common and 126 Arctic Tern nests were found. The timing of breeding was consistently later than at nearby colonies but breeding output was much higher than at Aberlady Bay. Possible explanations for the terns’ departure arn

re-establishment are discussed.

Introduction

Four species of terns bred on the Isle of May, Firth of Forth during the first half of the 19th century but the colony was aban- doned about 1850. Terns bred again from 1921 to 1957 with numbers varying greatly from year to year (details in Eggeling 1960). Common Terns Sterna hirundo were par- ticularly numerous in 1927 and 1936 and reached a peak of 5-6000 pairs in 1946-7; numbers of Arctic Terns S. paradisaea were highest in 1936 (800 pairs) and in 1946 (400-550 pairs) which was also the peak year for Sandwich Terns S. sandvicensis (1400-1500 pairs) and Roseate Terns S. dougallii (c.15-20 pairs). At its zenith in 1946 the colony held 6800-8100 pairs, but in subsequent years numbers of all species declined and breeding success was poor. For 24 years from 1958 there were only a few isolated breeding attempts, mostly by Common Terns (Eggeling 1974, Tasker 1979, Lack 1980, Bayes 1981). In 1982 a small colony of Common Terns was re- established and in 1984 Arctic Terns also resumed breeding. This paper documents these recolonisations.

Methods

To keep human disturbance to a minimum no detailed nest checks were made. However, each year visits of less than 20

minutes were made to the colony at approx- imately weekly intervals in good weather conditions. In 1982 nests were not marked but in 1983 clutches were numbered (felt tip pen on eggs). Since then all nests have been marked with short numbered stakes placed c.30 cm away from the nest. Occasional observations made from a temporary hide and/or the Lighthouse tower provided in- formation on the extent of the colony each year, the species present and their biology. The species count for each year was the cumulative total of nests found. A few of these nests were almost certainly repeats after earlier failures and so the counts presented will slightly overestimate the number of breeding pairs. The locations of place names mentioned in the text are shown in. Figs. 1.

Results Numbers and distribution

Between 1958 and 1981 there were seven nesting records of Common Terns and one of Arctic Terns (Table 1). Four pairs of Common Terns bred in 1981 but the nests were widely scattered both in space and time and it was not until 1982 that a well-defined colony was established. During May 1982 increasing numbers of Common Terns displayed over North Plateau just north of

2 S. Wanless

SB 15 (1)

250m

NORTH HORN

NORTH

PLATEAU THREE (~

TARN: NIG ae

LIGHTHOUSE>~\ SOUTH PLATEAU

WATER CATCHMENT

ory

. < a Sie 8 . pe . ve «os

RUFF GREEN

‘ ‘ ’ . : . on - . . . ws ss Ws eC = o = . ‘ ~. Oe i Ink ESR ome ‘. = . ' so4 vo 5 So .

Zf_ST ANDREWS WELL

a HIGH ROAD —— HOLYMANS

pS ee ROAD a a LOOKOUT

RE

ee

ws ‘ .

BEACON

FIGURE 1 Isle of May showing places mentioned in the text. The dotted lines indicate tracks.

the Lighthouse and two nests with single eggs were found in an area of sorrel Rumex acetosa on 27 May. Fourteen nests were found that year and the number of birds present reached a peak of 52 on 26 July (Table 2). In addition, one pair of Common Terns nested near the Water Catchment on South Plateau. The North Plateau area was re-occupied in 1983 and 1984 when respec- tively 37 and 36 nests with eggs were found. In 1984 seven pairs of Arctic Terns arrived with an influx of Common Terns between 15 and 23 June and subsequently 19 pairs of Arctic Terns bred in an area of close- cropped turf and rock overlooking the Common Tern colony. A pair of Common Terns probably attempted to breed on South Plateau.

In 1985 the two species showed the same segregation in the North Plateau col- ony as in 1984, apart from an isolated pair of Common Terns which nested below the Beacon. Numbers continued to increase and

80 Common and 87 Arctic Tern nests were found. In 1986, the Arctic Tern was the most numerous species with an estimated 128 nests; only 22 Common Terns nested and the colony was more fragmented than previously. Sixty-five clutches, mainly of Arctic Terns, were found in their usual area on North Plateau. There was a second more dispersed sub-colony of 85 nests between the Beacon and Holyman’s Road. At least 17 nests were Common Terns’ and most of these were either on the concrete base of the old Lookout or in the tall vegetation on the ash heap below the Beacon. Most of the Arctic Tern nests were on the short-cropped turf towards Holyman’s Road. Two pairs (? species) laid on the ridge running parallel to the High Road between the Lighthouse and Three Tarn Nick and another pair nested just south of St Andrew’s Well. No terns bred in the area originally colonised. On 22 July following a helicopter incident, the North Plateau colony was entirely

| |

1988

Terns on the Isle of May 3

TABLE 1 Details of breeding attempts by terns on the Isle of May, 1958-81. (Information taken from Eggeling 1974, Tasker 1979, Lack 1980 and Bayes 1981).

Year

1973 1979

1980

1981

Species of Tern Common Common

Arctic

Common

Common Common Common

Common

Location

Low Light - Tarbet

Water catchment, South Plateau

Water catchment, South Plateau

Water catchment, South Plateau

Centre of island, north of Low Light South Plateau

North Plateau

Comments

Nest with eggs, fate unknown.

Nest with one egg by 27 June. Failed.

Nest with two eggs on 5 July. Failed.

Nest with two eggs on 18 July. Hatched, young survived at least 6 days.

Nest with one egg in mid-June. Failed.

Nest with two eggs on 2 July. Failed.

Nest with one egg on 13 July. Hatched 3 Aug., fate unknown.

Nest with two eggs on 14 July. Hatched 5 Aug., young survived at least 9 days.

TABLE 2 Numbers of Common and Arctic Tern nests and hatching success (% eggs hatched of those laid) and overall breeding success (% young fledged from total eggs laid) on the Isle of May, 1982-7. (Information taken from Frazer 1982, Wanless 1983, 1984, 1986, Wanless & Ewins 1985, Dore & Wanless

1987).

Year

1982 1983 1984 1985 1986 1987

Number of nests

15 37 = 80 22 76

Tern - 15 = 37 19 56 87 167 128%, 2153* 126 202

Common Arctic Total Common Arctic Common Tern

Hatching success Overall breeding No. young No. young % success fledged fledged % per nest Arctic Species Species Tern Tern Tern Tern combined combined — — CxESF* — 7 0.47 36 ~ 13 - c.4-11 0.11—0.30 85 87.5 High High c.50-60 0.89—1.07 70 78 £7 Sal 8 0.05 c.95 c.95 Moderate® Moderate® c.80-100 0.52—0.65 High High Moderate Moderate c. 100 0.50

Notes: * includes unspecified nests.

OK

°

14 nests produced a minimum of 17 young of which at least seven fledged successfully. eggs and young from at least 25 nests were lost after disturbance by a helicopter.

4 S. Wanless

SB 15 (1)

deserted. As a result the eggs and small young from at least 25 nests were robbed by gulls; some of the larger chicks were pro- bably also killed although some apparently moved to the Beacon colony. Young which had already fledged did not return to be fed and apart from a few birds which flew over the colony in the days immediately follow- ing the incident, adults ignored the area for the rest of the season.

In 1987 the bulk of the terns (74 Com- mon and 123 Arctic Terns) nested within the Beacon area; also 2 pairs of Common and 2 pairs of Arctic Terns nested on North Plateau and one pair of Arctic Terns nested near St Andrew’s Well. It is possible that this shift in distribution was in response to the previous year’s helicopter disturbance.

In the 1930s and 1940s, the main breeding area was immediately below the North Horn, but during the1950s terns bred on the east side of Ruff Green, around St Andrew’s Well, the south end of North Plateau and the slopes between Holyman’s Road and the Lookout (Eggeling 1974). Recolonisation has thus occurred in all the latter areas except Ruff Green, which Eg- geling considered to be suboptimal, but the traditional stronghold at North Horn has not been recolonised.

Timing of breeding

Between 1973 and 1981 most clutches were not started until July (Table 1) but breeding was generally earlier once birds started to breed in groups. One pair of Common Terns had one egg by 16 May 1982 but lay- ing usually began at the end of May, with the main laying period from 10 to 25 June. The earliest chick hatched around 21 June and the first young fledged in the third week of July. The timing of laying was similar from year to year and was consistently later than at nearby colonies. For instance, from 1982 to 1987 the first Arctic Tern egg at Aberlady Bay was laid between 18 and 27 May and most Common and Arctic Tern clutches were initiated by the end of the first week of June (P.R. Gordon pers. comm.).

Similarly the first Arctic Tern egg on the Farne Islands was found each year around 16 May and the first Common Tern egg bet- ween 17 and 22 May (National Trust Wat- chers pers. comm.). At the Sands of Forvie comparable dates were 22-29 May for Arc- tic Tern and 22-31 May for Common Tern (A.J.M. Smith pers. comm.).

In many seabird species young birds tend to breed later than older ones (e.g. Coulson & White 1958). This effect has been found in Common Terns (Hays 1978, Nisbet et al. 1984), but there was no significant dif- ference in laying date amongst different age classes of Arctic Terns on the Farne Islands despite younger birds arriving consistently later (Coulson & Horobin 1976). However, the situation may well be different in a new- ly formed colony such as the Isle of May.

Breeding success

Four of the eight breeding attempts made between 1973 and 1981 failed soon after lay- ing, but one pair in 1981 possibly fledged two young (Table 1). Similarly, all isolated nests failed during the period 1982-7. In the main colonies, breeding success varied great- ly from year to year (Table 2). In 1985 ear- ly clutches were far more successful than late clutches: of clutches started before 25 June 81% (n= 58) of Common and 82% (n= 76) of Arctic Terns were successful, whereas comparable figures for those started on or after 25 June were 13% (n=18) and 0% (n= 10). In 1986 and 1987 there was no such obvious seasonal decline.

Causes of failure were mostly unknown but late clutches were sometimes abandoned and predation by Herring Gulls Larus argentatus and Lesser Black-backed Gulls L. fuscus of both eggs and chicks occurred in most years. For example, during 30 minutes on 16 July 1985, gulls made a total of 21 separate raids and at least two tern chicks were taken. Losses of both eggs and chicks tended to be greater when the weather was wet and windy; in 1987 at least 10 clut- ches were deserted and 69 dead chicks most- ly aged 1-2 days but a few up to 14 days old,

1988

Terns on the Isle of May 5

were found between 12 and 23 July during a period of persistent wet and windy weather. In part this may have been due to chilling but several studies have shown that high winds decrease the food capture rates by adult terns and the growth rate of chicks (e.g. Dunn 1975, Taylor 1983) so food shor- tage may have been a contributing factor.

Except for 1985, breeding success of Common and Arctic Terns on the Isle of May was similar to that found in previous studies (Langham 1972, Coulson & Horobin 1976, Hays 1978). It was markedly higher than the breeding output from the nearby colony at Aberlady Bay where a total of no more than 11 young Common/Arctic Terns fledged between 1982 and 1987 from an an- nual breeding population of 68-113 pairs i.e. a productivity range of 0-0.04 young/pair (P.R. Gordon, pers. comm.). Breeding suc- cess on the Isle of May was also generally higher than at the Sands of Forvie where,

except in 1987, output was less than 0.36 and 0.21 young/pair for Common and Arc- tic Terns respectively. The installation of an electric fence at Forvie in 1987 prevented predation by Foxes Vulpes vulpes and breeding success increased to 1.5 and 1.0 young/pair (details supplied by A.J.M. Smith). Terns occasionally do have breeding failures such as that observed on the Isle of May in 1985 (Bullock & Gommersall 1981).

Prey

In 1985 some observations were made on prey items brought in both for display by adults and to feed their chicks (Table 3). Sandeels Ammodytes sp. were the com- monest prey of both tern species. Clupeidea (mostly herring Clupea harengus) were caught more frequently by Common Terns but Arctic Terns took a wider variety of items. A similar range of prey types was recorded in the less detailed observations

TABLE 3 Prey items brought in by terns on the Isle of May, 1985. (Data from Wanless & Ewins 1985).

Common Tern

Prey Items only

Sandeels Ammoodytes sp. 27 26

Clupeidae (mostly herring Clupea harengus) 9 20

Three-spined stickleback Gasterosteus aculeatus

Cod Gadus morhua 0 Blenny sp. Blenniidae 1 Flatfish sp. 0 Unidentified fish 3 Prawn/Shrimp 1 Earthworm Lumbricus sp. 0 Unidentified 1

Totals 42 48

eye S&S & & © &

Display Chick Total Proportion food

DS

29

wo Oo-A wWwowoz- © &

Ne) oo

Arctic Tern

Display Chick Total Proportion

(%) of all only food (%) of all prey prey 54 26 100 126 66 27 10 25 nm. 35 18 8 0 0 0 0 0 0 1 1 0.5

1 0 0 0 0 0 5 5 3 10 0 8610 1 0 1 1 0.5 0 3 0 3 3 6 asf 55 137 292

Note: *items found lying in or near nests could not definitely be attributed to either display or

chick food categories.

T_T

6 S. Wanless

SB 15 (1)

made in other years, but rockling Ciliata sp. were taken in 1983 and 1984 and herring was thought to be the main prey item in 1984.

Evidence of prospecting by other tern species

Roseate and Sandwich Terns bred on the Isle of May between c.1930 and 1956 (Eg- geling 1960). There have been no records of the former showing interest in the island, but since 1980 Sandwich Terns have sometimes done so; a single bird was seen carrying fish on 28 July 1980 (Lack 1980), in 1985 c.10 birds regularly flew over the North Plateau in early June (one was seen carrying a display fish) and on 13 June some landed in the Common/Arctic colony (Wanless & Ewins 1985). On 18 June 1986 Sandwich Terns were seen displaying over South Plateau (Wanless 1986).

Discussion

In contrast to many seabird species whose numbers have increased in Scotland this cen- tury, terns are now generally less abundant than previously (Thom 1986). In a review of breeding terns in Britain and Ireland, Thomas (1982) identified some adverse fac- tors which have had marked effects on tern numbers between 1975 and 1979. These in- cluded high tides and bad weather, human disturbance, predation and gull problems. Eggeling (1960) considered that terns aban- doned the Isle of May in the 1950s because increasing numbers of Herring and Lesser Black-backed Gulls forced them to move from their favoured breeding area near the North Horn to the less favourable centre of the island. Gulls were seen eating tern eggs and chicks and, in the final years of the col- ony, the few terns which attempted to breed had extremely low nesting success. When terns finally deserted the island in 1958, there were c.3000 pairs of gulls (Eggeling 1960). Gull numbers continued to increase to c.17,500 pairs in 1972 (Eggeling 1974). At this time the Nature Conservancy started to cull breeding gulls on the island. Over the next ten years the gull population was reduc-

ed to 3-4000 pairs and by 1982, when Com- mon Terns returned, gull numbers were similar to those at the time of the desertion. Thus the terns’ absence from the Isle of May coincided with the period that the gull population exceeded 3-4000 pairs.

However, other factors could also be involved. For instance, young herring are often important prey for terns during the breeding season (Cramp 1985). The Firth of Forth once had a large spring-spawning her- ring stock which supported a considerable human fishery. Catches were good in the 1930s and early 1940s but fell rapidly from 1942 onwards, and between 1946 and 1963 virtually no adult herring were caught in the Forth (Saville 1963). Herring were apparent- ly still scarce in 1973 at the start of a long term study of the diet of young Puffins Fratercula arctica, and initially no herring of the size taken by terns were recorded (Hislop & Harris 1985). They first appeared in the Puffin’s diet in 1975 but it was not until 1980 that the proportion exceeded 10%. Since 1981 the proportion has rang- ed from 34-39% (Hislop & Harris 1985, M.P. Harris pers. comm.) and the years (1984, 1986) with the highest percentages of herring in the diet coincided with the years of highest breeding success for the terns. In the Firth of Clyde more than 50% of the annual variation in the number of Common and Arctic Terns breeding on Horse Island could be explained by changes in the abun- dance of post-larval and juvenile herring (Monaghan & Zonfrillo in press). Herring are obviously important in influencing the numbers of many species of seabird (e.g. Coulson & Thomas 1985, Aebischer 1986, Anker-Nilssen 1987). However, whilst there is broad agreement between the absence of terns from the Isle of May and apparently low herring stocks in the Firth of Forth, there are anomalies. First, the tern colony peaked in 1946 and 1947 just as the herring fishery collapsed. Second it persisted, albeit with reduced numbers of birds, for a fur- ther 10 years. Third, the terns’ recovery lagged behind the increase in herring abun- dance by several years.

1988

Terns on the Isle of May 7

Another factor could have played a part in the terns’ disappearance. In 1955 there was an outbreak of myxomatosis amongst the Rabbits Oryctolagus cuniculus on the Isle of May which was followed by marked changes in the vegetation during this and the next two seasons (Eggeling 1960). Although the main decline in tern numbers occurred before this, the long, rank vegeta- tion was possibly unsuitable nesting habitat, particularly for Arctic Terns which prefer close-cropped turf (Cramp 1985).

No natural mammalian predators (such as rats Rattus spp. or Foxes) have been recorded on the Isle of May but the Lighthouse keepers used to keep pets and in the early 1970s one dog was adept at dig- ging out and killing Puffins (M.P. Harris, pers. comm.). However, there is no men- tion in the Observatory records of dogs or cats affecting terns. Similarly it is hard to decide if human disturbance contributed to the terns’ desertion. Since their recolonisa- tion the terns have been protected by rop- ing off the breeding areas. In 1983 and 1984 decoys were put out to attract prospecting birds and in 1983 sound recordings of tern vocalizations were played. Such measures have been effective in re-establishing terns in other areas where they bred formerly (Kress 1983), and may have helped here.

The 1958-79 episode was not the first case of the island being abandoned by terns. During the first half of the 19th century there was a sizeable tern colony but this was deserted between 1850 and 1920 (Eggeling 1960). In this instance there were no large gulls nesting on the island (Eggeling 1960), the Forth herring fishery was thriving (Saville 1963) and myxomatosis was un- known (Matthews 1952). The true reason(s) for the terns’ desertion and recolonisation of the Isle of May is unlikely ever to be known but clearly there were several factors operating at the time which could have pro- duced the observed changes.

Acknowledgments

This paper draws heavily on observations made by NCC Isle of May summer wardens:— M.

Tasker, Dr. P. Lack, K. Bayes, M. Frazer, Dr. P.J. Ewins and C. Dore, and visitors to the Isle of May Bird Observatory. I am grateful to Dr. J.R.G. Hislop and D. McKay for information on fish numbers in the Firth of Forth and to P.R. Gordon, the Farne Island watchers and A.J.M. Smith for comparative data from Aberlady Bay, the Farne Islands and the Sands of Forvie. Thanks are also due to Dr. P.J. Ewins, Dr. M.P. Harris and P.K. Kinnear for improving the manuscript.

References

Aebischer, N.J. 1986. Retrospective investigation of an ecological disaster in the Shag Phala- crocorax aristotelis: A general method based on long term marking. J. Anim. Ecol. 55: 613-629.

Anker-Nilssen, T. 1987. The breeding perfor- mance of puffins Fratercula arctica on Rost, northern Norway 1979-85. Fauna noryv. Ser. C, Cinclus 10: 21-38.

Bayes, K. 1981. Isle of May NNR, Summer Warden’s Report. NCC unpub. report (South-east Region).

Bullock, I1.D. & Gomersall, C.H. 1981. The breeding populations of terns in Orkney and Shetland in 1980. Bird Study 28: 187-200.

Coulson, J.C. & Horobin, J. 1976. The influence of age on the breeding biology and survival of the Arctic Tern Sterna paradisaea. J. Zool Lond. 178: 247-260.

Coulson, J.C. & Thomas, C.S. 1985. Changes in biology of the Kittiwake Rissa tridactyla: a 31 year study of a breeding colony. J. Anim. Ecol. 54: 9-26.

Coulson, J.C. & White, E. 1958. The effect of age on the breeding biology of the Kittiwake Rissa tridactyla. Ibis 100: 40-51.

Cramp, S. (Ed.) 1985. The Birds of the Western Palearctic Vol. 1V. Oxford University Press, Oxford.

Dore, C.P. & Wanless, S. 1987. Isle of May NNR Summer Warden’s Report. NCC unpub. report (South-east Region).

Dunn, E.K. 1975. The role of environmental factors in the growth of tern chicks. J. Anim. Ecol. 44: 743-755.

Eggeling, W.J. 1960. The Isle of May. Oliver & Boyd, Edinburgh.

Eggeling, W.J. 1974. The birds of the Isle of May. Scott. Birds (supplement) 8: 93-148.

8 S. Wanless

Frazer, M.W. 1982. Isle of May NNR Summer Wardens’ Report. NCC unpub. report (South-east Region).

Hays, H. 1978. Timing and breeding success in three to seven year-old Common Terns. /bis. 120: 127-8.

Hislop, J.R.G. & Harris, M.P. 1985. Recent changes in the food of young Puffins Frater- cula arctica on the Isle of May in relation to fish stocks. [bis 127: 234-239.

Kress, S.W. 1983. The use of decoys, sound recordings and gull control for re-establishing a tern colony in Maine. Colonial Waterbirds 6: 185-196.

Lack, P.C. 1980. Isle of May NNR Summer Warden’s Report. NCC unpub. report (South-east Region).

Langham, N.P.E. 1972. Chick survival in terns (Sterna spp.) with particular reference to the Common Tern. J. Anim. Ecol. 41: 385-394.

Matthews, L.H. 1952. British Mammals. Collins, London.

Monaghan, P. & Zonfrillo, B. (in press). Popula- tion dynamics of seabirds in the Firth of Clyde. Proc. Royal Soc. Edinburgh.

SB 15 (1)

Nisbet, I.C.T., Winchell, J.M. & Heise, A.E. 1984. Influence of age on the breeding biology of common terns. Colonial Water- birds 7: 117-126.

Saville, A. 1963. A decline in the yield of Scottish estuarine spring spawning herring fisheries. Rapports et Procés Verbaux, Reun. Conseil Internat. pour L’exploration de la Mer 154: 215-219.

Tasker, M. 1979. Isle of May NNR, Summer Warden’s Report. NCC unpub. report (South-east Region).

Taylor, 1.R. 1983. Effect of wind on the forag- ing behaviour of Common and Sandwich Terns. Ornis. Scand. 14: 90-96.

Thom, V.M. 1986. Birds in Scotland. Poyser, Calton.

Thomas, G.J. 1982. Breeding terns in Britain and Ireland, 1975-79. Seabird Report 6: 59-69.

Wanless, S. 1983, 1984, 1986. Isle of May NNR Summer Warden’s Report. NCC unpub. reports (South-east Region).

Wanless, S. & Ewins, P.J. 1985. Isle of May NNR Summer Wardens’ Report. NCC unpub. report (South-east Region).

S. Wanless, Nature Conservancy Council, 46 Crossgate, Cupar, Fife KY15 5HS.

(Ms. received 30 November 1987)

ms chiar CRN eR ten MN

| Scottish Birds (1988) 15: 9-20

The distribution and status of Arctic and Great Skuas in Shetland 1985-86

P.J. EWINS, P.M. ELLIS, D.B. BIRD AND A. PRIOR

Totals of 1912 Arctic Skua and 5647 Great Skua Apparently Occupied Territories (AOTs) were recorded in Shetland in 1985-86, representing 61% and 76% of the British breeding population. On the Shetland Mainland Arctic Skua numbers have increased at c.4% p.a. and Great Skua numbers at c.9% p.a. since 1974-75. Human persecution and reclamation of moorlands may have affected numbers locally, but food availability was probably the main factor influencing skua numbers. Widespread breeding failure of terns in recent years has probably reduced Arctic Skua breeding output and

recruitment.

Introduction

Shetland and Orkney are the traditional British strongholds of the Arctic Skua Ster- corarius skua and the Great Skua Catharacta skua (Cramp & Simmons 1983, Thom 1986). Apart from early documentation of the Great Skua colonisation in the 19th century (e.g. Low 1879) no attempt was made to assess breeding numbers until the ‘Operation Seafarer’ surveys in 1969 and 1970 (Cramp et al. 1974). Unfortunately these surveys did not cover large areas of suitable skua breeding habitat inland and so the number of skuas was underestimated.

Much better coverage was achieved by the RSPB skua surveys in 1974 and 1975 (Everett 1982), but census techniques for breeding skuas have been greatly improved since then (Furness 1982). Using these refined census methods Meek ef al. (1985) conducted a complete survey of Orkney skuas in 1982, which revealed some dramatic increases. A full survey of breeding skuas in Shetland was clearly overdue.

There has been speculation that Arctic Skua numbers in Shetland may be declining, possibly due to reduced food availability stemming from poor breeding success of

Arctic Terns Sterna paradisaea and Kitti- wakes Rissa tridactyla in recent years (Furness et al. 1986, Heubeck & Ellis 1986). Following surveys in 1985 of moorland birds in several large areas of Shetland (Peacock etal. 1985, Wynde & Richardson 1985), we carried out surveys in 1986 to complete the coverage for skuas. The detailed results have been presented in a full report (Ewins et al. 1987). This paper summarizes the results and discusses population trends and factors influencing them.

Methods

We used the ‘Apparently Occupied Territory’

(AOT) as the count unit, as recommended by

Furness (1982). An AOT was scored for any of

the following:

(i) positive signs of breeding (e.g. nest, eggs or young)

(ii) incubating adult

(ili) distracting or alarming adult(s)

(iv) pair or single bird in potential breeding habitat and apparently attached to the area Skuas flying past, or feeding, or single birds

flushed from an area and which flew out of sight

were not recorded. Groups of three or more skuas

10 P.J. Ewins et al.

SB 15 (1)

seen together regularly but which showed no sign of territorial behaviour were recorded as ‘club’ (non-breeding) birds. Some failed pairs may have tolerated other birds in their territories, thereby causing us to underestimate AOTs.

Between May and July 1986 the NCC Upland Bird Survey carried out detailed surveys of terrestrial birds on 32 moorland plots in Shetland, using standard techniques described by Reed ef al. (1983). These surveys recorded the ac- tivity and position of every skua seen from the fixed transects. We translated this information in- to AOTs using the criteria outlined above. Counts of nests were used for Great Skuas on Noss (1983) and Hermaness (1985).

Moorland areas were divided into blocks which could be covered by one or two surveyors within a few hours. With the exception of the 1986 NCC plots the survey method involved walking transects up to a maximum of 500 m apart. Walk- ing along ridges was an effective means of cover- ing large areas in suitable locations. At the same

time, a watch was kept for skuas flying up ahead. In addition the observer stopped at regular inter- vals (c.200-300m) and scanned the area thoroughly from a suitable vantage point. In areas where skuas were numerous, scanning took longer and more transects were walked. In such colonies it was usually necessary to stand or sit for five minutes to allow birds to settle before meaningful plots of AOTs could be made. Two people walk- ing adjacent transects caused much confusion in high density areas, and we found that a single observer could census these areas more accurately by plotting AOT positions from various vantage points. Surveys were completed between 25 May and 8 July, the majority during June. This six-week period was selected on the basis of available in- formation on the timing of breeding in both species, and the timing of skua surveys elsewhere. There was no evidence that the timing of breeding in the two survey years differed markedly from normal (Ewins ef a/. 1987, Furness 1977a, Furness

TABLE 1 Shetland regional totals of skua Apparently Occupied Territories (AOT) and club birds, 1985-86. Roman numerals correspond to regions shown in Figs. 1 & 2. Proportions (%) are those in each

region of the Shetland AOT total.

ARCTIC SKUA GREAT SKUA Region AOT AOT Club AOT AOT Club n n n % n

| Unst 267 14.0 0 1257 22.3 203 I Yell 192 10.0 33 313 5.5 38 Hl Fetlar 180 9.4 0 248 4.4 35 IV Whalsay islands 41 2a 0 1 0 4 Vv North Mainland Was 9:2 0 23a 4.1 7 VI Central Mainland & 189 9.9 10 114 2.0

Yell Sound islands Vil West Mainland 169 8.8 0 96 Id 35 VIII South Mainland 326 17.0 6) 792 14.0 119

& islands IX Fair Isle 115 6.0 20 84 tS 54 X Foula 164 8.6 335 2495 44.2 858 Xl Papa Stour 94 4.9 0 14 0.2 0

TOTAL 1912 104 5647 1357

1988

Skuas in Shetland 1985-86 11

FIGURE 1 The number of Arctic Skua AOTs in different areas. Regions (Roman numerals) correspond with those in Table 1.

1980). Counts at either end of the day were avoid- ed, and were not carried out in winds exceeding Beaufort force 4, or in persistent rain, very cold conditions or fog.

The largest areas were covered in 1986 by the RSPB (Bird et a/. 1986), and the NCC. Many volunteers provided coverage of smaller areas, in conjunction with the Seabird Colony Register. Many areas of farmland and crofting in-bye were not surveyed because it was well-known that skuas were rarely seen on such habitat. Nevertheless, as a check various in-bye areas on the Mainland were observed from surrounding roads, and the anticipated absence of skuas was confirmed.

The colour phase of Arctic Skuas was deter- mined on the basis of belly colouration i.e. dark (melanic) or pale, as recommended by P. O’Donald (in Meek ef al. 1985).

Statistical tests are denoted by superior figures and the results are given in the Appendix.

-20 21-50 51-150

151-500

501-1000

~

-- _- -—--—

FOULA

XU

FAIR ISLE alle IX '@

FIGURE 2 The number of Great Skua AOTs in different areas. Regions (Roman numerals) correspond with those in Table 1.

Results

Totals, distribution and density

The surveys in 1985 and 1986 recorded 1912 Arc- tic Skua AOTs and 5647 Great Skua AOTs in Shetland, together with 104 club Arctic and 1357 club Great Skuas. Breeding skuas were found throughout Shetland but there was much regional variation (Table 1), which was not due simply to differences in the availability of nesting habitat. Breeding density also varied markedly between regions (see below).

Breeding Arctic and Great Skuas occurred most commonly in the northern isles (Unst, Fetlar, Yell), the south Mainland, and on some of the larger offshore islands. Arctic Skua AOTs were more regularly dispersed than Great Skua AOTs, although numbers were markedly lower in the central and east Mainland as far north as Ronas Hill (Fig. 1). Foula held nearly half of

12 P.J. Ewins et al.

SB 15 (1)

TABLE 2 Breeding number and densities of skuas in various breeding aggregations, 1985-86. Densities refer to the area occupied by the aggregation, not of the island or region.

ARCTIC SKUA GREAT SKUA

Location AOTs AOTs AOTs AOTs per per km2 km2

ISLANDS

Foula 164 133 2495 277

Fair Isle ql 85 81 45

Noss 15 150 388 242

Mousa Dal 53 9 30

Hascosay 10 33 29 26

Hermaness ‘tal 55 819 60

(Unst)

MAINLAND

North Roe 40 13 81 12

Mio Ness-Toft 31 42 10 6

Sandness Hill 20 17 31 11

W. Quarff 18 3 10 3

S. Mainland DA 12 57 10

hills (N)

S. Mainland 27, (6! 82 22

hills (S)

Shetland’s Great Skuas, and Unst just less than a quarter. Great Skuas were scarce throughout the west, central, east and much of the north Mainland (Fig. 2).

Breeding densities for each species were calculated for smaller islands, or discrete breeding aggregations on larger land masses by measuring the area of a polygon drawn around the outer- most AOTs of each breeding aggregation (but omitting isolated AOTs). There was considerable variation in breeding density throughout Shetland, with the highest densities of each species tending to occur on smaller islands and in the largest colonies (Table 2). This was not always the case though. A regression of Arctic Skua breeding density (Y) against the number of AOTs (X) of the adjacent Great Skua colony showed a highly significant linear relationship! indicating that in the vicinity of the larger Great Skua col- onies, Arctic Skuas bred at higher density. This applied both to colonies on islands and on the Mainland.

Characteristics of breeding territories

Distance from the nearest sea Due to the con- voluted nature of the coastline no part of Shetland is more than 5 km from the sea (Flinn 1974). Most skua AOTs were within 2 km of the sea (92% for Arctic Skua, and 97% for Great Skua). Skua con- centrations on small islands will obviously tend to over-represent the categories closest to the sea. However, even when the Foula data were omit- ted 95% of Great Skua AOTs were within 2 km of the sea. There was a highly significant dif- ference between the two species in the proportion of AOTs at different distances from the sea, both including? and excluding? the Foula Great Skua data. Overall, significantly more Great Skuas than Arctic Skuas bred within 1 km of the sea (79% vs. 60%)*.

Slope Slopes were measured from 1:25,000 OS maps, and AOTs categorised by slope of habitat i.e. shallow/flat (< 10°), medium (10-30°), or steep (<30°). On Foula all the Arctic Skuas bred on shallow/flat ground, but 36% of Great Skuas bred on slopes of medium gradient, and 9% on steep slopes. Ewins et al. (1986) showed than when compared with availability there was a significant tendency for Great Skuas to breed away from steeper slopes.

Elsewhere in Shetland no skua AOTs were found on ‘steep’ slopes, and ‘medium’ slopes sup- ported only 1% of Great Skua and 0.2% of Arc- tic Skua AOTs (these were all close to large seabird colonies).

Habitat type Most skua AOTs occurred on peatland dominated by heather Ca/luna vulgaris. The only area in which Great Skuas were found breeding on in-bye grassland was Foula; up to 10 pairs (0.4% of the Foula population) were involy- ed, and their breeding attempts were usually destroyed. Great Skuas preferred to nest amongst vegetation up to c.12 cm high, whereas Arctic Skuas often nested in shorter vegetation. Accor- dingly, Great Skuas were found nesting not only on peatlands dominated by heather, crowberry Empetrum nigrum and blaeberry/cowberry Vac- cinium myrtillus, V.vitis-idaea, but quite often in wetter flushes with larger stands and clumps of coarse grasses and rushes Juncus spp. Arctic Skua AOTs were found less frequently in such poorly- drained areas with taller vegetation, and most oc- curred on heathery bogs, or on drier moors dominated by very short heather and various herbs and lichens, often with substantial maritime influence. Moors overlying serpentine bedrock,

1988

Skuas in Shetland 1985-86 13

those on shallow soils, and those from which the peat had previously been removed by man, were used regularly by Arctic Skuas, but not by Great Skuas, probably because the vegetation was very short.

In a few areas Arctic Skuas had established AOTs on grassland. On Foula c.10 Arctic Skua AOTs (6% of the Foula total) were on in-bye grassland, and most bred, apparently successfully. On Fair Isle only 2 (2%) were on previously im- proved grassland, although a few pairs bred in mires dominated by rough grasses, where former drainage systems had not been maintained. The application of lime and surface seeding of the Eas Brecks area by the Bird Observatory on Fair Isle in 1984 has apparently not yet affected the number of Arctic Skua AOTs there. This flat area of short heather has traditionally been one of the core breeding areas on Fair Isle. By 1986 the alien grass mix had provided a green sward, but although most of the original heather was dead (due to the liming), the leafless, shrubby remains of the plants were presumably sufficient to en- courage the returning Arctic Skuas to establish AOTs here and breed. Elsewhere in Shetland Arc- tic Skuas bred only rarely on improved grassland, and we knew of no AOTs on areas re-seeded within the last decade.

Arctic Skua colour phases

The colour phases of birds in pairs and of single Arctic Skuas were pooled for regions with ade- quate sample sizes. This showed that the incidence of pale phase birds in the South Mainland, adja- cent islands and Fair Isle (regions VIII and IX in Table 3) was significantly lower than found fur- ther north and west in Shetland (18.3% and 23.8% respectively)>. In 1986 the overall in- cidence of pale phase birds in Shetland was 21.6% (Table 3), which was significantly lower® than the 26.5% quoted for Shetland between 1943 and 1979 by Meek ef al. (1985).

The colour phase of both members of 662 Pairings was recorded in 1986 (Table 4). There was a significant difference between the propor- tions in the S. Mainland, adjacent islands and Fair Isle, and those elsewhere in Shetland’. This was consistent with the previous result, that pale phase birds were more frequent further north in the Shetland archipelago. Similarly, the overall Shetland proportions had changed significantly from those recorded between 1943 and 19798. Overall, pale x pale pairings are now rarer, and melanic x melanic pairings commoner than previously (Table 4).

TABLE 3. The frequency of pale phase Arctic Skuas in 1986. Only those regions with reasonable sample sizes are detailed. N = total number of individuals whose colour phase was determined.

Pale phase

Region N n %

| Unst 188 41 21.8 lil Fetlar 184 43 23.4 V N. Mainland AT AD Zt VIC. Mainland etc. fisowst W 274 VIL W. Mainland TEA O28 25.9 VIII S. Mainland etc. 424 84 19.8 IX Fair Isle 226. 35 15.4 X Foula 23550055 23.4

Shetland 1986 1615 349 21.6

Discussion Comparison with the 1974-75 survey

Everett (1982) reported that the surveys in 1974 and 1975 found (estimated) at least 1631 pairs of Arctic Skua and at least 5451 pairs of Great Skua in Shetland. The surveys in 1985 and 1986 recorded 1912 Arctic Skua AOTs and 5647 Great Skua AOTs. A straightforward comparison of these totals suggests increases over the 11-year period of 17% for Arctic Skuas and 4% for Great Skuas, but such direct comparison of totals is invalid for a number of reasons:

(i) Coverage was incomplete in the

1974-75 surveys.

(ii) When examining original survey data/maps from 1974-75 we found some discrepancies between what was actually recorded, and the results sum- marised (and estimated) in Everett’s paper.

(iii) Estimates were made for some areas and islands in 1974-75 (Everett 1982). Our experience in 1986 has shown that, at least for some areas, local or- nithologists had a rather poor idea of skua breeding numbers.

14 P.J. Ewins et al.

SB 15 (1)

TABLE 4 The colour phases of Arctic Skua pairings in Shetland in 1986. p = pale, m = melanic, N = total number of pairings. Previous Shetland data from Meek et al. (1985).

PAIRINGS m x m x x eeu Region 0 NN ee | Unst 92. 60 65.2 24 26.0 8 8.6 Il Fetlar 92 55 59.8 31 33u7 6 6.5 VN. Mainland 16 11 68.8 4 25.0 1 6.3 VIC. Mainland etc. 66 35 53.0 30 45.5 1 1.5 Vil W. Mainland 41 DF 65.9 8 19.5 6 14.6 Vill S. Mainland etc. 150 96 64.0 53 35.5 1 0.7 IX Fair Isle 113 80 70.8 31 27.4 2 1.8 X Foula 89 52 58.4 29 32.6 8 9.0 Shetland 1986 662 418 63.1 210 ST. 34 5.1 Shetland 1943-79 376 218 58.0 120 31.9 38 10.1

(iv) The 1974-75 surveys used less rigorous census criteria, based on the scoring of “‘nairs holding territory’’. Bradley ef al. (1975) detailed what they regarded as a ‘pair’, and distinguished between ter- ritorial pairs, and non-breeding pairs, but they only regarded a single bird to be a breeder, or on territory, if it per- formed some distraction display. We are not aware that any consistent ap- proach was adopted in other areas then.

In attempting to compare results from the two surveys, we considered only those areas for which coverage was known to be complete in 1974-75. Although it is not known how the ‘pair’ census unit of 1974-75 would have compared with an AOT unit, for the purposes of this analysis we have regarded them as comparable.

The comparisons of totals between surveys for 12 islands and six areas on the Shetland Mainland revealed that for both skua species there was a good deal of varia- tion in the magnitude and direction of change between sites, although trends were apparent. For these sample areas (which bet- ween them held 75% of Shetland Arctic

Skuas, and 93% of the Great Skuas in 1985-86), the overall weighted mean increase was 0.8% for Arctic Skuas and 15.4% for Great Skuas. These figures represent average annual increases of 0.1% for Arc- tic Skua, and 1.3% for Great Skua.

For Arctic Skuas, the general trend was for a slight decrease on islands, but a substantial significant increase on the Mainland. Great Skua numbers had increas- ed significantly on islands, and even more so on the Mainland (Table 5).

Status and population trends

Using the most recently available informa- tion we have compared breeding numbers in different parts of the British range (no skuas are known to breed in Eire, N. Ireland, England or Wales). Although surveys in Orkney and Shetland have used the AOT as the count unit, counts and estimates elsewhere have usually been of pairs. We have not attempted to convert one unit to the other, and totals are given as either AOTs or pairs. Shetland supports 61% of Britain’s Arctic Skuas, with most of the remainder breeding in Orkney (33%). Great Skuas are even more restricted to

1988

Skuas in Shetland 1985-86 15

Shetland, with 76% of the British popula- tion breeding there, and 22% breeding in Orkney (Table 6).

The current Scottish breeding popula- tion of c.3150 pairs/AOTs of Arctic Skuas represents only 0.3-3% of the estimated world population, but in terms of the population in the NE Atlantic it is an important breeding concentration; the only larger populations are in Norway (8000 pairs) and Iceland (4000 pairs, Furness 1987).

Furness (1986) estimated that c.12,500 pairs of Great Skuas breed currently in the northern hemisphere. He used 5000 pairs as the Shetland total; this we can now increase to 5647 AOTs, making the best current estimate 13,147 pairs/AOTs. Therefore Iceland, with c.5000 pairs (38%), and Scotland with c.7425 AOTs (56%) are the strongholds, and within Scotland, Shetland alone holds 43% of the northern hemisphere population. The 2495 AOTs on Foula account for c.19% of the total northern hemisphere breeding population.

In the mid-1970s the Great Skua was considered to be increasing in Scotland at an average rate of 7% p.a. (Furness 1987).

A direct comparison of the 1974-75 survey total of 5970+ pairs (Everett 1982) with the 1985-86 figure (incorporating 1980s data for Orkney, Hebrides etc.) of c.7425 AOTs for Scotland suggests on average increase of only 2% p.a. over the last 11 years. Taking into account the improved coverage in recent surveys in the northern isles, and the small populations elsewhere, the increase of Scottish Great Skuas appears to be levelling off. A similar calculation for Scottish Arctic Skuas also reveals an average increase of 2% p.a. over the same period.

It is difficult to assess overall trends in Shetland precisely as survey methods and coverage have improved greatly in recent years, but different trends are evident on islands compared with the Mainland (Table 5). However, skua breeding numbers and their distribution have been recorded in more detail since 1960 at the major seabird stations of Fair Isle, Foula and Noss (Fig. 3). At these sites Arctic Skuas appear to have increased generally until the late 1970s then declined at varying rates into the mid-1980s. Great Skuas have increased dramatically here, though numbers on Foula appear to have stabilised or decreased

TABLE5 Comparison of skua totals in 1974-75 (pairs) and 1985-86 (AOTs) surveys, for groups of areas where coverage was complete in both surveys.

Number Number Weighted Mean % Level of pairs of AOTs mean change per of 1974-75 1985-86 % change annum change® ARCTIC SKUA Islands 1219 1115 —8.5 —0.7 ns Mainland 201 317 +57.7 +4.2 sgiel Total 1420 1432 +0.8 +0.1 ns GREAT SKUA Islands 4459 4905 + 10.0 +0.9 1 Mainland 184 351 + 145.1 +8.5 a

Total 4643 5256 + 15.4 +1.3 te

Note: ° Gross changes compared statistically with the null hypothesis that no change had occurred, using G tests. ** P< 0.01; *** P<0.001; ns non-significant.

16 P.J. Ewins et al.

SB 15 (1)

TABLE6 Most recently available counts of breeding skuas in Scotland. (SBR = Scottish Bird Report).

ARCTIC SKUA ~=GREAT SKUA Count No. % of No. % of Year unit AOTs total AOTs total Source

SHETLAND

Foula 1986 AOT 164 5.2 2495 33.6 Ewins et al. (1986)

Fair Isle 1986 AOT MS 34 84 1,1, _ FIBOTGinstitt)

rest Shetland 1985-86 AOT 1633 51.8 3068 41.4 This report ORKNEY 1982 AOT 1034 32.8 1652 22.2 Meek et al. (1985) OUTER HEBRIDES

Lewis 1982 pair 20-30 0.8 20 0.3 SBR

N Uist 1984 pair 52 7 0 0 SBR

Benbecula 1982 pair 6-10 0.2 0 0 SBR

St Kilda 1984 pair 0 0 37-40 0.5: SBR

N Rona 1982 pair 0 0 6 0.1 Furness (1986)

Shiants 1977 pair 0 0 2 0.03 SBR ARGYLL ISLANDS

Coll 1985 pair 30 1.0 0 0 SBR

Jura 1985 pair 10 0.3 0 O SBR

Colonsay 1975 pair 1 0 0 0 Everett (1982) SUTHERLAND

Handa 1985 pair 35 led 52 0.7" "SBE

Mainland 1980-85 pair 1-3 0.1 0 0 Thom (1986)

Summer Isles 1982 pair 0 0 2 0.03 Furness (1986)

Eilean Roan 1981 pair 0 0 1 0.05 SBR CAITHNESS 1979-80 pair 40+ 1.3 2 0.03 Thom (1986),

Reed et al. (1983) Minimum 3146 7424

TOTNES OTS tee tna 3160+ 7427

slightly since a peak in the mid-1970s. Possible explanations for some of these trends are discussed below.

Factors affecting skua numbers in Shetland Food Supply

The distribution of both skua species corres- ponds fairly well to that of the main seabird colonies (see Figure 26 in Berry & Johnston 1980). The more even breeding dispersion of Arctic Skuas almost certainly reflects their greater dependence on Arctic Terns for obtaining food through kleptoparasitism. Tern colonies are quite evenly spread through Shetland (Bullock & Gomersall

1980). Although studies of Arctic Skua diet near auk breeding concentrations have found piracy of auks to be a major source of food (Furness 1980, Ewins 1986), our observations indicate that piracy of Arctic Terns is the main source of food in other areas.

During the 1970s and early 1980s sandeels (probably mostly Ammodytes marinus) formed the base of the Shetland seabird food pyramid (Furness 1977a, 1980, Furness & Hislop 1981, Ewins 1985, 1986, M. Heubeck pers. comm.), associated with population increases of many seabird species (M. Heubeck pers. comm.). Over the last

1988 Skuas in Shetland 1985-86 17 ARCTIC SKUA GREAT SKUA 3000 f% a =o 300 2000 yf FOULA VA 1000 WA 0 200 : ROUEA” “800 F » i NOSS (aa / : 300 / 2 pi 100 FAIR = -200F »-------- tees 100 FAIR NOSS [Ee 0 (i ee ee 0 1960 70 80 86 1960 70 80 86

| four to five years however, sandeels seem to have become less readily available, par- ticularly to the surface feeding seabirds, and very poor breeding seasons have resulted (Heubeck & Ellis 1986, Furness et a/. 1986, Coulson & Megson 1987). Total landings of the Shetland sandeel fishery, and the catch per unit effort, have also been declining steadily since reaching a peak in the early 1980s (Warburton 1983, R. Bailey pers. comm.).

Great Skuas are very catholic in diet and feeding habits, and although adults are opportunistic feeders they raise their young mainly on sandeels which they catch themselves. Piracy appears to be a fairly unimportant feeding method, employed when preferred feeding conditions are poor (Furness 1987). At present, Great Skuas have alternative food sources such as offal, whitefish discards, and eggs, young and

FIGURE 3 Changes in numbers of breeding Arctic and Great Skuas on three Shetland islands since 1960.

adults of almost any bird species. In some parts of Shetland a few Great Skuas have specialised in taking eggs and live chicks, fledglings and adults of a variety of seabirds (Andersson 1976, Ewins 1985, Furness 1987, Harvey & Suddaby 1986, M. Heubeck pers. comm.). The forthcoming increase in the minimum mesh size for UK trawlers has raised concern that the resultant decrease in availability of whitefish discards (of suitable size to be available to Great Skuas), com- bined with the decline in sandeel availabili- ty closer inshore, may force Great Skuas to increase their exploitation of breeding seabirds (Furness 1987). Early suggestions that this swing has already started were evi- dent in 1986 at a number of Kittiwake col- onies (M. Heubeck & P. Harvey pers. comm.). Such a change in diet was found in Brown Skuas Catharacta lonnbergi nesting on Rangatira, Chatham Islands

18 P.J. Ewins et al.

SB 15 (1)

(New Zealand): following the removal of the island’s sheep, skua numbers halved and they switched from sheep carrion to heavy predation on nocturnal petrels (Young 1978).

Breeding habitat

Shortage of suitable nesting habitat largely explains the relative scarcity of Great Skuas near seabird colonies on some parts of the Mainland and on many smaller islands. Agricultural reclamation of heather moorland has invariably resulted in skuas moving to nearby nesting habitat. Such loss of nesting habitat need not in itself change skua breeding success or numbers, but there is presumably a limit beyond which nesting territories cannot decrease in size. Great Skuas could in theory breed at much higher densities than at present in most areas (ex- cluding Foula and Noss), and we suspect that food availability rather than breeding habitat is the main factor limiting their numbers in most parts of Shetland. However, breeding success of Great Skuas is lower at high densities, and so reduced recruitment in dense colonies may be effec- ting a density-dependent control on numbers (Furness 1984). The breeding suc- cess of Arctic Skuas was lower at high nesting density, but not significantly so (Furness 1980).

Inter-specific competition for limited breeding habitat has occurred in some areas, and since the Great Skua arrives back on the breeding grounds first, and is a much stronger, dominant species, it is the Arctic Skua which has been forced to modify its breeding dispersion in response to increas- ing Great Skua numbers. On Unst, Fair Isle, and some parts of the Mainland Arctic Skuas have been forced out of certain areas, but with no significant overall decline in numbers. On Noss however, Arctic Skua numbers have dropped and the density in breeding colonies increased, largely due to displacement by Great Skuas. Arctic Skua hatching success there was 65% in 1974, compared with 80% in 1946. This was associated with an increase in breeding den-

sity from 62 to 157 pairs/km? over the period, and an increase in egg losses to other Arctic Skuas (Perry 1948, Kinnear 1974). Fledging success was thought to be similar in the two studies, and Great Skuas were never seen to take Arctic Skua eggs, and they seldom killed chicks (Furness 1977b). The situation on Foula is more com- plex because although Arctic Skuas were similarly forced into small pockets of breeding habitat by the expanding Great Skua population, numbers actually increas- ed at the same time as breeding density. The population increase was due to improved feeding conditions (via escalating Arctic Tern numbers and presumably greater availability of sandeels from incoming auks and Kittiwakes). Arctic Skua numbers then decreased in parallel with tern numbers (Furness 1977b, Ewins eft al. 1986). Thus food availability rather than breeding habitat is probably the main factor influen- cing Arctic Skua, as well as Great Skua breeding numbers in Shetland at present. In Orkney, moorland habitats are scarcer than in Shetland, and Meek ef al. (1985) suspected that Arctic Skua population size in some areas was actually limited by availability of breeding habitat.

Great Skua predation on Arctic Skuas

Various authors have claimed that the in- creases in Great Skua numbers this century have caused declines in Arctic Skua colonies in Shetland, but this has been based only on circumstantial evidence (Furness 1977b). However, Great Skua predation of Arctic Skua fledglings is now regularly observed on Noss (McKay & Crossthwaite 1985, Harvey & Suddaby 1986). In 1985 a max- imum of 32 Arctic Skua chicks fledged on Noss, of which between 12 and 22 were kill- ed subsequently by Great Skuas (a post- fledging mortality of 38-69%). Unlike the situation on Foula, this heavy Great Skua predation and ‘squeezing’ of Arctic Skua territories appears to be having a con- siderable effect, and if it continues Arctic Skuas may be unable to remain as a breeding species on the island.

1988

Skuas in Shetland 1985-86 19

Away from the smaller islands, this inter-specific predation (which is usually of fledglings but may also involve adults) may occur at low levels, but as Arctic Skuas can usually find alternative breeding habitat at a ‘safer’ distance from the high density Great Skua areas, they trade increased distance to food supplies against improved breeding success and adult survival.

Persecution

There is no evidence that either recruitment to, or the size of, Shetland’s Arctic Skua population has been significantly affected through the limited persecution in a few areas. The much heavier, though still localis- ed, persecution of Great Skuas (mainly of adults) has in the past been a key factor con- trolling numbers both locally and in Shetland as a whole. Today it is still impor- tant at the local level (Furness 1987). However, ringing recoveries indicate con- siderable interchange between breeding areas for both species in Shetland (though not of established breeders), and gaps created by persecution, particularly in favoured areas, are quickly filled by immigrants.

Great Skuas are prevented by man from colonising certain areas, but when this persecution is relaxed the species spreads rapidly into suitable breeding areas. On Fair Isle this has been at the expense of Arctic Skuas which have been forced to breed at higher density in adjacent areas. Therefore in such areas, human persecution of Great Skuas (on account of their alleged preda- tion on lambs) has favoured Arctic Skuas by maintaining preferred breeding habitat free from Great Skuas.

Acknowledgments

These surveys would not have been possible without the combined funding of the NCC, RSPB, Seabird Group, and Shell UK Ltd. We thank these organisations and in particular the following individuals who helped with the plan- ning and administration work: Drs L. Campbell,

R.W. Furness, M. Pienkowski and M.G. Richardson, and R. Grant, M. Heubeck, E.R. Meek, J. Reid and M.L. Tasker. The coverage achieved was partly due to the volunteers, whose efforts are gratefully acknowledged. Draft manuscripts were improved by Drs L. Campbell, R.W. Furness, A.V. Hudson, M.G. Richardson and D. Carstairs. Finally, many thanks to all the Shetlanders who allowed us to wander so freely over their moorlands.

References

Anderson, M. 1976. Predation and kleptopara- sitism in a Shetland seabird colony. Jbis 118: 208-217.

Berry, R.J. & Johnston, J.L. 1980. The Natural History of Shetland. Collins, London. Bird, D.R., Prior, A., Ellis, P.M. & Ewins, P.J. 1986. RSPB Skua Surveys in Shetland 1986.

RSPB unpubl. report.

Bradley, S., Fisher, P. & Round, P. 1975. A census of moorland birds in Shetland, June/ July 1975. NCC unpubl. report.

Bullock, I1.D. & Gomersall, C.H. 1981. The breeding populations of terns in Orkney and Shetland in 1980. Bird Study 28: 187-200.

Coulson, J.C. & Megson, G. 1987. The popula- tion and breeding biology of the Arctic Tern Sterna paradisaea in Shetland, 1986. NCC unpubl. report.

Cramp, S. & Simmons, K.E.L. (Eds). 1983. The Birds of the Western Palearctic. Vol. III. Oxford University Press, Oxford.

Cramp, S., Bourne, W.R.P. & Saunders, D. 1974. The Seabirds of Britain and Ireland. Collins, London.

Everett, M.J. 1982. Breeding Great and Arctic Skuas in Scotland in 1974-75. Seabird 6: 50-59.

Ewins, P.J. 1985. Growth, diet and mortality of Arctic Tern Sterna paradisaea chicks in Shetland. Seabird 8: 59-68.

Ewins, P.J. 1986. The ecology of Black Guille- mots Cepphus grylle in Shetland. D Phil thesis, University of Oxford.

Ewins, P.J., Wynde, R.M. & Richardson, M.G. 1986. The 1986 census of Arctic and Great Skuas on Foula, Shetland. NCC Unpublished report.

Ewins, P.J., Bird, D.R., Ellis, P.M. & Prior, A. 1987. The distribution and status of Arctic and Great Skuas in Shetland, 1985-86. Un- published report to NCC, RSPB, Seabird Group and Shell UK Ltd.

20 ~=P.J. Ewins et al.

SB 15 (1)

Flinn, D. 1974. The coastline of Shetland. Pp. 13-23 in: Goodier, R. (Ed). The Natural Environment of Shetland. NCC, Edinburgh.

Furness, B.L. 1980. Territoriality and feeding behaviour in the Arctic Skua Stercorarius parasiticus (L.). PhD thesis, University of Aberdeen.

Furness, R.W. 1977a. Studies on the breeding biology and population dynamics of the Great Skua Catharacta skua (Brunnich). PhD thesis, University of Durham.

Furness, R.W. 1977b. Effects of Great Skuas on Arctic Skuas in Shetland. Br. Birds 70: 96-107.

Furness, R.W. 1981. The impact of predation by Great Skuas Catharacta skua on other seabird populations at a Shetland colony. bis 123: 534-539.

Furness, R.W. 1982. Methods used to census skua colonies. Seabird 6: 44-47.

Furness, R.W. 1984. Influences of adult age and experience, nest location, clutch size and lay- ing sequence on the breeding success of the Great Skua. J. Zool., Lond. 202: 565-576.

Furness, R.W. 1986. The conservation of Arctic and Great Skuas and their impact on agriculture. NCC unpubl. report.

Furness, R.W. 1987. The Skuas. Poyser, Calton.

Furness, R.W. & Hislop, J.R.G. 1981. Diets and feeding ecology of Great Skuas during the breeding season in Shetland. J. Zool., Lond. 195: 1-23.

Furness, R.W., Harris, M.P., Hunt, J., Riddiford N. & Wanless. S. 1986. The seabird scene. Scott. Bird News 4: 4-5.

Harvey, P.V. & Suddaby, D. 1986. Wardens’ final report for 1986, Noss NNR. NCC un- publ. report.

Heubeck, M. & Ellis, P.M. 1986. Shetland sea- birds 1985. BTO News 143: 10.

Kinnear, P.K. 1974. Report on skua survey carried out on Noss NNR, 1974. NCC un- publ. report.

Low, G. 1879. Orkney and Schetland. Facsimile reprint, 1978. Melven, Inverness.

McKay, C. & Crossthwaite, S.K. 1985. Noss summer wardens’ report for 1985. NCC un- publ. report.

Meek, E.R., Booth, C.J., Reynolds, P. & Ribbands, B. 1985. Breeding skuas in Orkney. Seabird 8: 21-33.

Peacock, M.A., Beveridge, F.M. & Campbell, L.H. 1985. Shetland moorland breeding birds survey 1985. RSPB unpubl. report.

Perry, R. 1948. Shetland Sanctuary. Faber & Faber, London.

Reed, T., Langslow, D.R. & Symonds, F.L. 1983. The breeding waders of the Caithness Flows. Scott. Birds 12: 180-186.

Sokal, R.R. & Rohlf, F.J. 1969. Biometry. Freeman, San Francisco.

Thom, V.M. 1986. Birds in Scotland. Poyser, Calton.

Warburton, 1983. Sandeels - the elusive species. Scott. Fish. Bull. 47: 22-27.

Wynde, R.M. & Richardson, M.G. 1986. Moor- land and coastal bird surveys - NCC 1985. NCC unpubl. report.

Young, E.C. 1978. Behavioural ecology of lonnbergi skuas in relation to environment on the Chatham Islands, New Zealand. N.Z.J. of Zool. 5: 401-416.

APPENDIX. Results of statistical tests

L... r=0.66, 21d.f.,. P< Ce Regression equation Y=0.05X + 21.2 2. G=272, 3d.f., P< 0.001 Gor Ges see Sokal & Rohlf 1969) G=80, 3d.f., P< 0.001 G=259, Idi.) P< 02001 G=6.94, ld:f:;, P< 0.01 G=8.3, 1d.f.P<0:01 G=8.82. 2d.f... P02 G—9 2. 2d.f. PaO

On NUN WwW

P.J. Ewins, NCC, Archway House, 7 Eastcheap, Letchworth, Herts SG6 3DG P.M. Ellis & D.R. Bird, RSPB, Seaview, Sandwick, Shetland

A. Prior, Happy Hansel, Walls, Shetland

(Revised ms. received 8 December 1987)

Scottish Birds (1988) 15: 21-29

21

The seabirds of St Kilda, 1987

MARK L. TASKER, PETER R. MOORE

AND RICHARD A. SCHOFIELD

A count of the seabirds on St Kilda in 1987 revealed that around 400,000 pairs breed within the island group.

The colonies of Puffin, Gannet and Leach’s Petrel are important on a world basis, while those of Fulmar, Storm Petrel, Kittiwake, Guillemot and Razorbill are important within Europe. The populations are at present healthy and the greatest future threat is considered to be the

accidental introduction of Brown Rats.

Introduction

St Kilda, situated 57°49’ N 08°35’ W, 66km WNW of North Uist, Outer Hebrides, holds western Europe’s most important seabird colony. The St Kilda group consists of one main island (Hirta), garrisoned by the Ar- my, three outlying uninhabited islands (Dun, Soay and Boreray) and severa! sea stacs (Fig. 1). All the main islands and Stac an Armin and Stac Lee hold large numbers of seabirds. A full description of the geography and habitats on the islands may be found in Williamson & Boyd (1960) and Jewell et a/. (1974).

Before the evacuation of the indigenous human population in 1930, seabirds were ex- ploited as a source of feathers, food and fuel. Martin (1698) mentioned the large numbers of birds, and subsequent records up to 1978 have been collated by Harris & Murray (1978).

Counts of the seabirds at St Kilda are difficult because much of the coastline is in- accessible and several cliffs exceed 450 m. The first attempt at a comprehensive count of all species was made for ‘‘Operation Seafarer’? (Cramp et ai. 1974), although there had been several earlier counts of single species (e.g. Boyd 1960). Harris & Murray (1978) repeated the total colony cen- sus in 1977, and summarized all previous counts. Duncan ef a/. (1982) counted the birds on Boreray in 1980. We visited the

islands in June 1987 to survey the popula- tions for the Seabird Colony Register organised by NCC and the Seabird Group.

Methods

Most of the study was done from 12-24 June 1987. Observers were present on Hirta throughout that period, on Boreray from 13-20 June, on Dun from 15-20 June and on Stac an Armin on 16 June. Where possible, the counts were made from land. Cliff sections that were not visible from safe vantage points were counted in calm weather from 20 m and 26 m converted fishing vessels or in- flatable craft. All counts were made in good visibility with no precipitation and in winds less than Force 4 (Beaufort Scale). Standard Seabird Colony Register count units (given in Results for each species) were used. Counts of Guillemots (Scientific names are given in the species sum- maries) and Razorbills were made between 0700 and 1500 GMT. The counts of gull territories were possibly made too late in the year, as many nests were empty and some adults might have left their territories. This might have been the consequence of a poor breeding season. Counts of Gannet nests were not attempted in 1987.

Burrow and hole nesting birds are more dif- ficult to census. No attempt was made to count Manx Shearwaters, Storm Petrels or Leach’s Petrels, but the distribution of calling birds was mapped by walking across parts of the islands bet- ween 2230 and 0230 GMT (Leaper et al. in press). Although close to mid-summer, most nights were overcast and dark and there was little moonlight.

siencicraegter

22 MLL. Tasker et al. SB 15 (1)

STAC

AN ARMIN it

Leacan an BORERAY t’Sluic Mhoir

Mor Conachair

Cleitan

Mol Ghiasgar McPhaidean

HIRTA

Mullach Bi Oiseval

Carn Mor

1 Village Bay

Caolas an Dun

FIGURE 1 Hirta, Soay and Dun, with inset of Boreray. Place names mentioned in text and location of sections used in Table 2 are shown.

Calling Storm Petrels were heard on several nights from 2200 GMT. A small sample count of oc- cupied Leach’s Petrel burrows was made on Boreray both by playing a tape of the call and sniffing the entrance of suitable holes. A burrow was considered to be inhabited if a bird was heard or there was a strong smell. All Black Guillemot sightings, whether on land or at sea, were noted and the counts presented comprise the sum of the highest counts made during the period in various parts of the archipelago. This is not the ideal time or method to count Black Guillemots; better and more repeatable counts may be obtained early in the morning during April (Ewins 1985) and it is likely that our counts of this species were underestimates.

The number of occupied Puffin burrows on the south slope of Boreray (Fig. 2) was assessed using a series of 30 m2 quadrats positioned at random in the colony. A line of bamboo canes 50 m apart was placed across the slope of the col- ony about 270 m above sea level. The upper and lower limits of the burrows were measured from this line. The lower limit was commonly the boun- dary between turf and bedrock. The colony was

then mapped on squared paper with each square representing a 100 m2 (10 m x 10m).

FIGURE 2 The extent of Puffin colonies on Boreray, 1987.

1988

Seabirds of St Kilda, 1987 23

These squares were numbered consecutively. Within the map of the colony, 34 points were selected using a table of random numbers. These points were then located on the ground and, using a piece of string 3.08 m long for the radius, the number of occupied burrows in a circle of 30 m2 was counted. The south slope colony was then divided into two sections, one dense, the other less so. The area of these sections was calculated from the colony map, and the mean number of occupied burrows in the 30 m? circles in each section was multiplied by the area of the section to produce an estimate of occupied bur- rows. These estimates were summed to give a total estimate for the colony.

A similar procedure has been used in previous years in the Puffin colony on Dun (Har- ris & Rothery in press). M.P. Harris and MLT counted occupied burrows in the random monitoring scheme on Dun on 21 and 22 May 1987. Sampling was not carried out at the other Puffin colonies (i.e. the Sunadel and west slope colonies Boreray, Soay, Hirta, Stac an Armin). The areas of Sunadel and the west slope colonies on Boreray were however assessed by eye, and an estimate for numbers of burrows in these areas was obtained from the mean burrow density of the less dense section of the south slope colony.

Results and comments on status

Fulmar Fulmarus glacialis A total of 62,786 apparently occupied Fulmar nest-sites was counted (Table 1). Over half the sites were on Hirta, the coastline of which was divid- ed into 10 sections (Table 2, Fig. 1) for com- parison with previous counts. The total sug- gests an increase of about 43% since 1977 both on Hirta and overall (Table 2). In 1977 (and on most previous occasions) counts were made in July (Harris & Murray 1978), so it is difficult to tell how much of the ap- parent increase might be attributable to the timing of the counts. A few Fulmars now nest inland on Hirta; this habit started around 1981 (W. Wright pers. comm.). Elsewhere there were obvious extensions in distribution when compared with that recorded on sketch maps in 1977 (S. Mur- ray in litt.). If all of the increase was ge- nuine, we estimate that the colony must have been increasing at a mean rate of 3.6% per annum, which is well within the capabili- ty of the species. Fisher (1952), for instance,

found the colony had grown from 20,780 nests in 1939 to 38,178 nests in 1949.

Manx Shearwater Puffinus puffinus The only known nesting sites in the group are on Hirta, Dun and Soay; no new nesting sites were located on these islands (Soay was not visited). No birds were found during searches of Boreray by night, although not all suitable habitat was visited.

Storm Petrel Hydrobates pelagicus On Boreray, calling birds were found around Cleitan McPhaidein and along the main southern ridge. On Dun, the species was confined to the rocky sections of the ridge. Similar rocky cliff edges and ridges were oc- cupied on Hirta and in addition some man- made structures were used. In general, Storm Petrels appeared to prefer structures with walls more than 1 m thick situated on dry ground. Thinner walls and structures on wet ground were not used.

Leach’s Petrel Oceanodroma leucorhoa On Boreray, all Leach’s Petrel burrows ap- peared to be situated within the Puffin col- ony on the south side up to 440 m above sea level. Burrows had not previously been found above 200 m (Duncan ef al. 1982). A total of between four and eight nests was found in an area of 100 m2 of the Puffin colony that was searched for nests. If this density estimate was representative of the remainder of the south slope colony (total area 79,400 m2) then the site holds approx- imately 3200-6400 occupied burrows. On Hirta, Leach’s Petrels were located above the cliffs of Oiseval, in the boulder field on the west of the Cambir, in Carn Mor and along the cliffs at Mullach Bi, as well as in small numbers at several locations. Calling birds were found throughout the steeper grass slopes on the north-west end of Dun. There have been no previous estimates of numbers on St Kilda.

Gannet Sula bassana In 1985, Murray & Wanless (1986) found 50,050 sites (Table 1). As there were no obvious changes in the ex- tent of the colony, the species was not counted in 1987.

SB 15 (1)

24 MLL. Tasker et al.

LOSOEC ZL VL8E SOLCC 6C8Z 9S

6S vSL vS

cS 0S00S

wr we Oe

98209

Je}O]

LBET UIWUY ue DP}S

LCSEL 0

0

0

6£

397] 2e1S

000€9 0 L OOOSLL O09LP c s 6 0 Z OL ZSZ 6 g €97 608L 6Z9€ 97 ZPZL 6L7Z ss BH9Z €767Z 0 6Z 90EL LEZL cL OL 0 c ZL SE € 0 0 v ZL 0 0 0 EL z 0 0 g 0 Z 0 0 v LZ 9/9b7Z 0 0 0 0 d j 0 i d d j 0 j d j 0 0 j d Z089 0g Zeb 6Z9§ _— SLOZL soel\s

Aeiaiog ysiuade] punos Aeog - Aeos ung pue|s|

‘smoling paidn390 = qo ‘sjenpiAipul = = }0e ‘aduasqe/aduasaid = e/d ‘sayis paidnd50 Ajjuaiedde = soe *Zg6l ‘epliy 3S JO Sde}S puke Spue|s! UO Spiig JO SJaquINU [2}0] | FIAVL

soe

us)

UIyJNd JOWII|IND 49k |g [|!qQuozey yoway|INDZ dYeMIVIY

HIND 9-g yeeIH IND Bulsapy HIND G-g Jesse] eNyS }ea1H 3eys

(S86L) JouURDH J21}9dg S,YyoeaT j91}9q WI0}S JayemMieays XUeW Jewyn4

sainads

SAPUI ‘SAlJO}119} paidnd90 Ajjuasedde

Seabirds of St Kilda, 1987 25

1988

GOLZZ GR0ZZ 0ss0z OLOSL 678Z 9r8S G8rLL OZZZ 98229 LL6EV IWLOL 97 0€ 0Z Ov 0 0 0 0 08 ZL YSIUdA2T 9EbL ELEL ogeL 07Z 9ZE L9z 02S OZ LBEZ 0002 UIWUY UP 2e}S 06 00€ 09b 00z cvz 8SL OLE 002 6€ 0S 32] 2k1S Zyl cSaL OOZL 0897 6Z €g 08 OSL Zeb 99¢ s9e}1S punos Aeros 8197 908 Ost 07 LEZL O€bL GZZL OZEL SLOZL 0r69 jeJO] 697% ZOLE OZOL 06L 6€0L SPZL ce0L OrLL Z 6ZE tb9 Og€ O€ Z6L Z8L 06L OgL L unq 6Z9E 9661 O9LE OLLZ €76Z 6SZL 09ZE O€EZ Z089 00SE [IO] O£9L 8ZS OOZL OSrL 0691 666 0zSZ OOZL ZLbZ € rrOL 06Z 098 OL9 rhb4 r9OL OL 09€ 196 Z COOL 879 009 OLL LL 96S OE OLZ 9ZE | Aeisai0g 6L7Z cca O€9L 0092 90EL ZL6 O%bzZ Oz9OL 6296 00€9 jeIO] PLLL G7Z 0S9 028 6SZ 9672 ose 008 pil € 7Z9 0S6 00Z OrSL G8S OLZ OOLL OL OZEZ Z €€p OgL 08Z 061 Z9p Ore WA Olt GE6L L Aeos S9rOL O€60L OOZOL 08€9 6LZL €ZZL OZOE 0802 6PESE 608bZ jeJO] Pv 0 Aeg ase\iiA 091 Ma} Bow ua] LU 89 0z 0OL ZL 0 0 0OL 7Sb OL 88h 887 O€€ 00S 0 0 0 0S ZO6L 6 LS6 66EL Ore 006 Ob GLZ 08z O0€ LLOv g 086 ZELL 09ZL 09rL Geb pee 0€6 O€S LLZb Z €€S 6€b 08 OOL 9rL 9rL OrL 0ZL 0z9 9 956 C18 OLZ 029 rel LL 08L 0Z OSL c vZLp 69S 096 OZ6L 9SE TPE O8LL 0€Z 66LLL p 8E9L cOzL 068 OL9 9ZL 8 O€Z OrL 778 € OL OL OLL 0Z 0 0 0S 0z p8ZL Z Z 0 0 0 OOL L6 08 0z 029 [ werewie) Z86L ZL6L 696L 6S6L Z86L ZL6L 696L 6S6L Z861 ZL6L yOWs]IND ayeMIIy Jewyjn4 uOI}DaS puR|s]

"SUOIIDAS PUPS! JY} 1Oj ajQqej!eAe JOU SI G96 PUe GG6L Ul SHeWYNY 104 UOIJEWJOJU “| “BI4y Bas ‘(Q96L) pAOg Aq padanins asoy) aie Aeos pure Aesasog ‘eWIH JO SUOI}DAS AY) “(APNYs SI4y}) Z861L pue (8Z6L Aeunw 8 SIdeH) ZZ6L “CH Ul pAOg "W'() 696L ‘“(CO96L PAO) 6SEL U! EP]Iy 1S UO SJowATIND [eNpIAIpul puke s}sau ayeMI}}1y ‘SA}IS AeWU]NY JO SJUNOD Z ITGVL

26 MLL. Tasker et al.

SB 13{7)

Shag Phalacrocorax aristotelis The 52 Shag nests found in the group represent a minimum number breeding on the islands, as nests are difficult to locate. There were two main groups: at least 20 at the Fort on Dun and 21 in Mol Ghiasgar, Hirta. A minimum of 70 adult birds was seen at the Fort and as about 90 pairs were present here in 1984 (Moore 1984), we feel that our count may have underestimated the size of this col- ony. The Mol Shoay boulder field on Soay had at least four nests, but no search was made on land; 32 adults were counted stan- ding on rocks. The two nests on Boreray are considerably less than the 50 pairs estimated to be present by Duncan ef al. (1982). We think there is insufficient suitable habitat (e.g. large crevices and boulder fields close to the sea) on Boreray for this number of pairs. Over 200 nests were estimated on St Kilda in the mid-1970s (Harris & Murray 1978).

Great Skua Stercorarius skua A total of 54 apparently occupied territories was found (Table 1). Of the 44 on Hirta, 35 were in Gleann Mor and nine between Conachair and Mullach Mor. Eight birds were on Soay and there were two territories on Boreray, where no nests were found in 1977 (Harris & Murray 1978) but one pair was present in 1980 (Duncan ef a/. 1982). Twenty-five pairs were present on Hirta in 1978 (Harris & Murray 1978).

Common Gull Larus canus This species bred in Glenn Mor in 1986. No nests were found in 1987, but a pair showed territorial behaviour at the north-west end of the Gleann.

Lesser Black-backed Gull Larus fuscus A total of 154 nests or territories was found (Table 1). The largest concentration of nests was in Gleann Mor; 38 were in the steep rocky area at the north-west end, 28 on the rocks of Leacan an t-Sluic Mhoir, and a fur- ther 15 on rocks at the head of Glen Bay. Eighteen territories were found on the Cam- bir, and 12 and 13 territories on Boreray and

Dun respectively. These figures represent apparent decreases since the mid-1970s, when there were between 160 and 240 pairs on Hirta and 30 nests on Dun (Harris & Murray 1978). It is possible however, that the count may have been after the main nesting season. Four nests were recorded on Boreray in 1980 (Duncan ef al. 1982). Har- ris & Murray (1978) recorded this species on Soay but we did not see it there.

Herring Gull Larus argentatus The majori- ty of the 59 apparently occupied Herring Gull territories found were on Boreray (Table 1). Harris & Murray (1978) did not record any on Boreray, but Duncan et al. (1982) found 49 nests in 1980. Since 1974 there has been an apparent decline from 40 to 14 pairs on Hirta and from 24 to 4 pairs on Dun. These counts may have been after the main nesting season.

Great Black-backed Gull Larus marinus Fifty-six pairs were located. Levenish had the densest colony with 10 pairs on its top. The 15 pairs found on Boreray represent a decrease since 1980 when around 30 pairs were estimated to be on the island (Duncan et al. 1982). Numbers on Hirta and Dun have declined since the mid-1970s from 31 and 40 pairs (Harris & Murray 1978) to 13 and 12 pairs respectively.

Kittiwake Rissa tridactyla Over one-third of the 7829 nests were on Boreray (Table 1), the main colonies being on the west side. Comparison with earlier counts (Table 2) reveals that the population is very similar to that recorded in 1959, but in the interven- ing period numbers apparently fluctuated.

Guillemot Uria aalge Although the total Guillemot population has remained cons- tant since 1969 (Table 2), there have been changes in numbers on each of the islands. On Boreray numbers appear to have in- creased by around 80%, while on Dun they have declined (Table 2). The colonies at the base of Conachair are the most important, holding 46% of the total.

ia

1988

Seabirds of St Kilda, 1987 27

Razorbill A/ca torda A total of 3814 in- dividual Razorbills was counted throughout the islands (Table 1). The largest concentra- tion, at the Fort on Dun, held over 1000 in- dividuals which is similar to the number counted in 1977. Other concentrations of over 100 birds were found in the boulder fields at Mol Ghiasgar on Hirta, Mol Shoay on Soay and on the eastern shoulder of Stac an Armin. Counts have been poor in the past, so it is difficult to assess any changes.

~ Black Guillemot Cepphus grylle Seventeen

birds were recorded on the water around the islands in much the same locations as before. There are two or three pairs on Caolas an Dun and two birds were observ- ed off Gob a Ghaill, Soay.

Puffin Fratercula arctica There were 29,600 (s.d. 3300) occupied burrows in the part of Dun we monitored. As this section held an estimated 71% of the island’s population in 1975, there may now be about 41,600 occupied burrows on Dun. There appears to have been little change since 1975 when the island was estimated to have 40,097 occupied burrows (Harris & Murray 1977).

On Boreray the denser section of the colony on the south slope had an area of 29,975 m2. The mean density of occupied burrows was 11.7 (s.d. 7.2) per 30 m2, giving a mean estimate of 11,637 occupied burrows for the section. The less dense sec- tion had an area of 49,450 m2 and a mean occupied burrow density of 5.4 (s.d. 5.0) per 30 m?, giving a mean estimate of 8819 occupied burrows. The combined area of the two other main colonies on Boreray was assessed as 238,500 m2; using the mean density estimate for the less dense section of the south slope, these sections of the island probably hold about 42,500 occupied burrows. The total number of occupied Puf- fin burrows on Boreray is thus about 63,000. This figure is lower than Brooke’s (1972) estimate of 77,000 occupied burrows, and Harris & Murray’s (1977) estimate of 100,000 pairs.

Puffins were found under many of the

cleits (stone shelters) on Stac an Armin and we estimate about 100 pairs there. The col- ony on Soay has long been considered the largest on St Kilda, with estimates ranging from 80,000 occupied burrows in 1971 (Brooke 1972) to 150,000 pairs in 1977 (Har- ris & Murray 1978). We did not survey the island but used the mean (115,000 pairs) of these estimates. The population of Hirta was not examined in 1987 but Harris & Mur- ray (1978) estimated 8100-13,500 pairs (mid- point 10,800). These figures give a grand total for the whole group in excess of 230,000 occupied burrows.

Discussion The importance of St Kilda

St Kilda is an internationally important site for breeding seabirds. It holds about 20% of the North Atlantic population of Gan- net (Murray & Wanless 1986), around 20% of the world population of the grabae race of Puffin (Harris 1984) and the largest col- ony of Leach’s Petrel in the eastern Atlan- tic. The Fulmar colony is the largest in Western Europe, although there are greater numbers at sites further north in the Atlan- tic. Over 1% of the British population of several other species breed on the islands. The Storm Petrel colony is probably the largest in Britain. Up-to-date figures for the total British populations of Guillemot, Razorbill and Kittiwake have yet to be calculated, but it is likely that the numbers of these species on St Kilda represent over 1% of the total European Community populations. These figures justify the status of St Kilda as a National Nature Reserve, its recent designation as a World Heritage Site and its proposed status as a Special Pro- tection Area of the European Community.

Status of the seabird populations

Our counts show that most of the popula- tions of seabirds on St Kilda are either similar to those of 10 years ago, or have in- creased. It is difficult to detect any trends in the Puffin population as Soay, on which

28 M.L. Tasker et al.

SB 15 (1)

50% of the estimated total breeds, was not surveyed. The number of large gulls on the islands has declined as have some popula- tions elsewhere in the British Isles (Lloyd & North 1987).

Conservation

The greatest potential threat to the bird populations on the islands would be the ac- cidental introduction of rats or other ground predators. As Harris & Murray (1978) pointed out, it is surprising that rats have not already arrived on Hirta. We consider that current precautions against the in- troduction of rats are not adequate. The military base on Hirta is supplied by regular visits of a landing-craft that runs onto the beach in Village Bay. This landing craft visits sites in the Clyde and the Outer Hebrides to collect supplies for St Kilda, and large numbers of Brown Rats Rattus norvegicus have been observed on the docksides in both of these areas. We con- sider that urgent action should be taken to reduce the possibility of rats reaching the islands.

An oil pollution incident near the islands during the breeding season (February to October) could be very serious. A load- ed tanker lost all power and drifted within 2 km of Boreray in December 1981. The re- routing of tankers that previously used the Minch to a shipping lane to the west of the Outer Hebrides, has undoubtedly increas- ed the pollution risks to St Kilda. The threat is greater than might at first appear, as studies at sea in summer 1987 (Leaper et al. in press) have shown that the most impor- tant feeding area for auks during the breeding season lies near the recommended deep-water shipping lane to the east of the islands. We would not however suggest that shipping be routed through the Minch, as this area holds important concentrations of birds throughout the year (Benn ef al. in press), whereas the waters to the west of the Outer Hebrides are used by large numbers of auks and other vulnerable seabirds only during the breeding season.

Acknowledgments

These counts were carried out as part of NCC’s Seabirds at Sea Project sponsored by the Depart- ment of Transport (Marine Pollution Control Unit), the Department of Energy, the Department of the Environment for Northern Ireland, Shell U.K. Ltd., BP Petroleum Development (UK) Ltd., Esso Petroleum Co. Ltd., Chevron Petroleum (UK) Ltd., Hydrocarbons Great Bri- tain Ltd. and NCC. The counts were carried out by the authors and by S. Benn, K. Camphuysen, M. Harman, M.P. Harris, G. Leaper, D. Miller, H. Prendergast, D. Thompson, P. Walsh and A. Webb. Some counts were made from MVs ‘‘Ocean Bounty”’ and ‘‘Charna’’; we thank the captains, Alistair Simpson and Stuart Moir and their crews for their seamanship. David Miller and the Army detachment on St Kilda gave us logistical help on the islands. Dr M.P. Harris and S. Murray supplied unpublished information. This paper was improved by comments from Drs M.P. Harris, C.S. Lloyd, M.W. Pienkowski and the editorial panel.

References

Benn, S., Burton, C.A., Tasker, M.L., Webb, A. & Ward, R.M. in press. Seabird distribution on the north-west Scottish shelf. NCC Chief Scientist Directorate report No. 803.

Boyd, J.M. 1960. The distribution and numbers of Kittiwakes and Guillemots at St Kilda. Br. Birds 53: 252-264.

Brooke, M.L. 1972. Population estimates for Puffins on Soay and Boreray and assessment of the rate of predation by gulls. Brathay Field Studies Report 20: 4-6.

Cramp, S., Bourne, W.R.P. & Saunders, D. 1974. The seabirds of Britain and Ireland. Collins, London.

Duncan, N., Taylor, K., Wanless, S., & Wood, V. 1982. The birds of Boreray, St Kilda. Seabird Report 6: 18-25.

Ewins, P.J. 1985. Colony attendance and censusing of Black Guillemots Cepphus grylle on Shetland. Bird Study 32: 176-185.

Fisher, J. 1952. The Fulmar. Collins, London.

Harris, M.P. 1984. The Puffin. Poyser, Calton.

Harris, M.P. & Murray, S. 1977. Puffins on St Kilda. Br. Birds 70: 50-65.

Harris, M.P. & Murray, S. 1978. Birds of St Kilda ITE, Cambridge.

1988 Seabirds of St Kilda, 1987 29

Harris, M.P. & Rothery, P. in press. Monitoring Lloyd, C.S. & North, S.G. 1987. Seabirds of of Puffin burrows on Dun, St Kilda, 1977-87. Troup and Pennan Heads, 1979-1986. Scott. Bird Study Birds 14: 191-198.

1 Jewell, P.A., Milner, C. & Boyd, J.M. (Eds). Martin, M. 1698. A late voyage to St Kilda. 1974. Island survivors: the ecology of the Brown & Goodwin, London. Soay sheep of St Kilda. Athlone Press, Moore, P.M. 1984. St Kilda warden’s report, London. 1984. NCC unpubl. report (NW Scotland

Leaper, G.M., Webb, A., Benn, S., Prendergast region).

H.D.V., Schofield, R.A., Moore, P.M. & Murray, S. & Wanless, S. 1986. The status of the Tasker, M.L. in press. Seabird studies at St. Gannet in Scotland 1984-85. Scott. Birds 14: Kilda, June 1987. NCC Chief Scientist Direc- 74-85.

torate report No. 804. Williamson, K. & Boyd, J.M. 1960. St Kilda

summer. Hutchinson, London.

Mark L. Tasker, Peter R. Moore, Richard A. Schofield, Seabirds Team, Chief Scientist Directorate, Nature Conservancy Council, 17 Rubislaw Terrace, Aberdeen ABI 1XE

(Revised ms. received 3 December 1987)

. - ‘ Z " - e | * ° : + * " 2 . Ss

Puffins on Dun K. Taylor

30 §©Scottish Birds (1988) 15: 30-35

Greenland White-fronted Geese in Caithness

S. LAYBOURNE AND A.D. FOX

Numbers of Greenland White-fronted Geese wintering in Caithness have declined from over 730 in the 1970s to c.220 in spring 1987, although numbers in the county remain of international importance. While the reasons for the decline remain unknown, continued loss of traditional peatland habitat has led to increasing

numbers resorting to agricultural land.

Introduction

The world population of Greenland White- fronted Geese Anser albifrons flavirostris breeds in west Greenland and winters exclusively in Ireland, northwest Scotland and two sites in Wales. Ruttledge & Ogilvie (1979) estimated a total of 14,300-16,600 birds in the late 1970s, which was a marked decline from the 17,500-23,000 estimated in the 1950s. The first simultaneous census of mainland Britain, in November 1982, gave a total of 7,189 (Stroud 1984a). This had risen to 10,858 by November 1986 (Stroud 1986), probably due to protection through- out much of its summer and winter range and a very good breeding season in 1985. Simultaneous counts on mainland Britain and in Ireland (organised by the Department of Tourism, Fisheries and Forestry, Dublin) gave a grand total of 22,353 Greenland Whitefronts in November 1985 (Norriss & Wilson 1986).

While an increase in the population is encouraging, results giving only totals mask the continued decline of many of the smaller, remote flocks in Scotland and Ireland away from the main wintering grounds of Wexford Slobs in south-east Ireland and Islay in the Inner Hebrides.

The peatland roost sites of Greenland White-fronted Geese in Scotland and Ireland are now under threat from commer- cial peat-cutting and forestry activities. The internationally important numbers of Whitefronts in Caithness have fluctuated

over the years and little was known of their flock structure, feeding or roosting sites. For this reason, efforts have been made since the late 1970s to obtain accurate counts of Whitefronts in Caithness and to examine the relationships between their feeding areas and roost sites. In this paper, we document results of the counts in Caithness over this period, and of a more intensive study carried out in March 1985.

Methods

Regular counts of White-fronted Geese have been carried out in Caithness over a number of years as part of the annual grey goose census organised by the Wildfowl Trust. However, no attempt had been made to count at all the known wintering sites of Greenland Whitefronts, or to arrange simultaneous counts of geese at these sites. Since 1981/82, SL has carried out detailed counts of Whitefronts to coincide with international counts throughout Britain and Ireland. All the Caithness resorts have been visited during a single day in the autumn and spring census periods (mid-November and late March/early April respectively), and additional counts have been made through the winter.

To eliminate the possibilities of inter- change between flocks during his day-long counts, intensive observations were carried out by a team of 12 observers using four cars in Caithness between 24 March and 3 April

1988

Greenland Whitefronts in Caithness 31

FIGURE 1. Map of Caithness showing geographical areas described in the analysis of Greenland White- fronted Goose use.

1985 (Laybourne & Fox 1985). All areas ever known to attract Greenland Whitefronts for feeding or roosting were visited at least once during the survey. All Whitefront observations were recorded and plotted with details of land-use on 1:10,000 OS maps. Simultaneous counts of all areas found to hold geese during the early part of the census were carried out on 29 and 30 March.

In addition, a review of earlier counts was undertaken, using the National Wild- fowl Count database and the literature sources referred to below.

Results

Maximum counts of Greenland White- fronted Geese during November gave totals

for the county ranging from 278-735 bet- ween 1970 and 1980; more than 350 were present in all years except 1970. However, some counts are maximum, unco-ordinated counts from different sites and it is difficult to assess the reliability of these totals. There seems little doubt that throughout that decade there were at least 350-400 birds pre- sent, but the absolute numbers are unknown.

Since 1981, better count coverage has been possible with all sites visited during one day. The spring count totals are shown in Table 1. These improved counting techni- ques have shown that there are five main groups of White-fronted Geese in Caithness (see Fig. 1 for locations).

32. S. Laybourne & A.D. Fox

SB 15 (1)

TABLE 1 Co-ordinated spring counts of Greenland White-fronted Geese in Caithness, 1981/82 —

1986/87.

Lochs Lochs

Calder Scarmclate

and Watten &

Year Broubster Winless 1981/82 180 TE 1982/83 246 77 1983/84 140 67 1984/85 124 233 1985/86 15% 54 1986/87 80 0

1. The Westfield Group

This discrete group utilises agricultural land to the south-west of Thurso. Numbers throughout any one season are consistent and up to nine colour-ringed birds caught in Greenland during 1979 have been present here but not elsewhere in Caithness (Labourne 1987). The flock feeds in the valley of the Forss Water in cereal fields and permanent pasture on the deeper fertile soils, rough pasture higher up and on some peaty accumulations on the hilltops. The geese start feeding on spilled grain on stub- ble, but transfer to pasture later in the season when this source of food is ex- hausted. They regularly roost on Broubster Leans, a complex valley bog with a range of wetland communities and open water, but will resort to Loch Calder to the east if disturbed.

Three hundred and seventy birds were counted in this group during 1978/79, but this had fallen to 220-250 by 1979/80, and to 80 in March 1986 (Table 1). There has been much reclamation of rough pasture and fields have been drained and reseeded and are no longer used by geese. In spring, Greylag Geese Anser anser are increasingly using the areas of wetter rough pasture preferred by the Whitefronts. The reseeded leys are also very attractive to Greylags and licences have been granted to disturb them. This undoubtedly also caused disturbance to Whitefronts at a time when the geese are

Total Loch Other Spring Heilen Areas Count 42 0 299 160 (52 635 120 41 368 2 60 419 25 0 226 144 0 224

rapidly accumulating reserves for the spring migration.

2. The Loch Heilen Flock

Loch Heilen is set in rich pasture on wind- blown sandy soils and is used by the geese for roosting and feeding. This flock ranges widely north to fields along the coastal strip between St John’s Loch and the Loch of Mey, which the flock uses as an alternate roost despite disturbance there from duck shooting. These birds are known also to feed on the in-bye fields of old crofts in the ex- tensive peatlands to the south and east of Loch Heilen. In the early 1970s, this group regularly flighted to and from Loch Meadie which was being used as an alternative roost, although they no longer do so. There may be some interchange between this flock and that of the peatlands considered below, but two colour-ringed birds consistently reported from the Heilen flock have not been seen in other flocks so any interchange is probably minimal.

In 1979/80, 190 Whitefronts were present in the Heilen area, but just over 144 were counted in 1986/87. There are several hundred Greylags near the loch in spring and autumn, and licences have been granted to disturb them. Many of the Whitefront feeding areas are remote, and there is an abundance of alternative feeding sites in the vicinity, so it is likely that these birds are not too disturbed.

]

)

1988

Greenland Whitefronts in Caithness 33

3. Central Group

Loch Scarmclate is an important site for wintering wildfowl and the gently sloping areas of reseeded and rough pasture by the loch are much frequented by Greylag and White-fronted Geese. Whitefront numbers have remained between 100 and 150 between 1979/80 and 1985/86; they use stubble fields north of Scarmclate early in the season and move to permanent pasture later. In the past, these geese have been seen flighting between here and Loch Heilen at dusk, but in recent years have only used Loch Scarm- clate. The Whitefronts also feed south and east to Loch Watten and further south to the marginal reclaimed land on the edge of the extensive areas of flow country further south to Loch of Toftingall, where birds are known to have roosted in the past. During the winters of 1985/86 and 1986/87, this flock apparently deserted the immediate area Of Loch Scarmclate but may have joined those roosting at Loch Heilen or Loch Meadie.

Anglers use Loch Scarmclate, and some adjacent areas have applied for licences to shoot Greylags. There has been relatively little agricultural change in this area, although the blanket mire about Loch of Toftingall has been afforested in recent years.

4. Moss of Killimster Flock

The Moss of Killimster is an area of species-rich lowland blanket mire with well- developed surface patterning, which is the traditional habitat of the Greenland Whitefront. This area with the adjacent swampy vegetation of Loch of Winless, wet meadows and reseeded fields supports a flock of c.40 geese which seems discrete. Examination of past counts from this area also suggests that 40-50 birds have been consistently present here. Geese have been seen feeding on stubble fields north of the loch and a group of c.40 feeding on pasture adjacent to Loch of Wester was presumably this group.

The presence of this flock was con-

firmed in 1984/85. It was previously assumed to be part of the Loch Heilen flock to the north. However, the presence of c.40 birds in the area since the early 1970s suggests this small group may have been in this part of Caithness for some time. Disturbance to the Killimster area is slight, with safe and quiet areas available nearby. The Loch of Wester area is subject to goose shooting to disturb Greylags and a trackway manufacturing lengths of continuously welded pipeline also detracts from the site. Upstream of the Loch of Wester, there has been extensive afforestation of lowland blanket mire which was probably important for the geese in the past.

5. South-western Flock

The Loch More area of Caithness has held Whitefronts since the end of the last century, when Harvie-Brown & Buckiey (1887) found the keeper at Strathmore bred pinioned birds descended from slightly wounded geese shot on the peatlands about the Lodge. Loch Meadie and Loch a’Cherigal were roost sites throughout the 1970s since twice-daily flights occurred between here and Loch Heilen. This roost flight ceased as winter progressed and has not been noted with any regularity in recent years. A flock of 60 flew in to roost on 27 March 1985, and birds were heard on other evenings at this site, but where these birds were feeding remains unknown. The up-rooted stems of cotton grass Eriophorum angustifolium with their bases eaten were found at a peatland flow site in 1985/86 and 1986/87, proving the birds were using the blanket mire regularly if only for night-time roost feeding.

Discussion

The northern and western distribution of wintering Greenland White-fronted Geese in Britain is associated with that of oceanic blanket bog with ‘“‘pool and hummock”’ sur- face topography. Here, the over-wintering geese eat the subterranean parts of the

34 S. Laybourne & A.D. Fox

SB 15 (1)

white-beaked sedge Rhynchospora alba and cotton grass which are common in such mires. There are few ground frosts to deny the geese this food and the soft Sphagnum carpets where the plants grow remain accessible in most winters.

The continuing loss of traditional peatland wintering habitat is a major cause for concern. With improving drainage techniques and ever more efficient machinery, the wetter, more heavily surface- patterned mires are now under as much threat for peat extraction, agricultural improvement and afforestation as the most easily drained bogs. Ruttledge & Ogilvie (1979) ascribed the decline of Greenland Whitefronts at 35% of sites and the deser- tion of 61% of sites to these factors. The destruction of Irish bogs continues (Ryan & Cross 1984) and drainage of the most important roost site in Britain for these geese began at Eilean na Muich Dubh SSSI (Duich Moss) on Islay in August 1985, although this has since ceased (see Stroud 1984b, 1985). At the time of writing, the flocks of Greenland Whitefronts at Loch Shiel (Lochaber), Moine Mhor (Argyll), Loch Sgoud (Wester Ross) and some of the Caithness group are the only ones using traditional boglands for substantial daytime feeding. The numbers involved amount to less than 0.1% of the total Scottish winter- ing population. With the exception of the Brent Goose Branta bernicla feeding on intertidal vegetation, the mire-feeding Greenland Whitefront now represents the last European goose species still wintering on natural (rather than agricultural or semi- natural) habitat. With such small numbers involved and the continued loss of this specialised habitat, the future for Greenland White-fronted Geese feeding on traditional areas looks uncertain.

In the past, Greenland Whitefronts have shown an inability to utilise alternative habitats when bog feeding has been denied them. The most dramatic example of this was the extinction of 600 birds in central Wales due to the hard weather of 1962/63 (Fox & Stroud 1986). However, at many

wintering sites, flocks now feed on farmland, particularly the poorer, wetter pasture, eating the roots and tubers of plants such as buttercups Ranunculus bulbosus, R. flammula, marsh arrow grass Triglochin palustre and lady’s smock Cardamine pratensis (D.A. Stroud pers. comm.). Even in these situations, the geese still often use peatlands as roost sites and may feed on Eriophorum during the night (as at Duich Moss, Stroud 1984b, 1985). While geese in Caithness are increasingly using agricultural areas, their roosting sites remain lochs and peatland lochans which are also used to a greater or lesser extent as feeding areas. The peatland roost sites are now under threat from commercial peat cutting and afforestation (Stroud ef a/. 1987). Although at present there are few conflicts between Greenland Whitefronts and agriculture, it is clear that continued loss of peatland habitat will result in an increasing proportion of the population resorting to farmland.

White-fronted Geese were described as ‘‘not an uncommon species in Caithness’’ at the end of the last century (Harvie-Brown & Buckley 1887), but Berry (1939) states “‘none of the observers in this area makes any mention of the White-fronted Goose, nor was the writer able to hear of any wintering in Caithness in 1931-32’’. Little is known of their subsequent status until the early 1960s when they were confirmed as belonging to the Greenland race. Atkinson- Willes (1963) showed peak counts of 10-100 for Caithness. Perhaps as many as 730 Whitefronts were present in the 1970s, but the county total, which numbered 635 in spring 1983, has since declined to 224 in spring 1987. Even allowing for this decline, the county total is still of national importance.

The decline of the Caithness Whitefront flocks over recent years gives cause for concern. Although most of their roosting sites are now protected as SSSIs, their feeding areas vary from year to year (depending on land-use management). The relationship between the geese and the peatlands is poorly understood and needs

ene

1988

Greenland Whitefronts in Caithness 35

further study before this feeding habit is lost forever. On agricultural land, protection from disturbance in late spring when they are feeding prior to leaving for the breeding grounds would seem advantageous. The provision of refuge areas would also reduce the current high level of disturbance on much of the farmland. Such refuges, managed specifically for Barnacle Geese Branta leucopsis have been very successful at Caerlaverock on the Solway (Owen et al. 1987) and at Loch Gruinart on Islay, where the less intensively managed grasslands have also proved effective as feeding refuges for Greenland Whitefronts.

Conservation, however, is concerned with the maintenance of regional natural diversity and cannot be restricted to the establishment of nature reserves. The Greenland White-fronted Goose has a long association with Caithness and now poses a unique conservation problem; continued monitoring of the population and protection of the remaining peatlands is required to protect the birds for the future.

Acknowledgments

The annual winter census of Greenland White- fronted Goose Study in Great Britain is supported by the NCC. The census of Caithness in 1985 received additional funding from the SOC, RSPB and Wildfowl Trust. Our thanks go to J. Bratton, J. Claricoates, D. Gilbert, M. Green, J. Hesp, C. MacKay, C. Mitchell, J. Moore, N. Penford and S. Ridgill for carrying out the census, often in appalling weather. Special thanks go to C. MacKay for logistics and to Dr. T. Keatinge (NCC) for help and advice. D. Stroud co-ordinated the work, provided historical information and generally conceived the whole project. Thanks also to N. Picozzi for considerable improvements to the manuscript. Finally our thanks to all those landowners who granted access to their land, and took the trouble to talk geese with us.

References

Atkinson-Willes, G.L. 1963. Wildfowl in Great Britain. Nature Conservancy Monograph 3. HMSO, London.

Berry, J. 1939. The Status and Distribution of Wild Geese and Wild Duck in Scotland. Cambridge University Press, Cambridge.

Fox, A.D. & Stroud, D.A. 1986. The Greenland White-fronted Goose in Wales. Nature in Wales (New Series) 4: 20-27.

Harvie-Brown, J.A. & Buckley, T.E. 1887. A Vertebrate Fauna of Sutherland, Caithness and West Cromarty. Oliver & Boyd, Edinburgh.

Laybourne, S. 1987. Greenland White-fronted Geese in Caithness. Caithness Bird Report (1986).

Laybourne, S. & Fox, A.D. (Eds). 1985. Report of the Caithness Greenland White-fronted Goose Spring Census 1985. Unpublished report, GWGS/NCC Inverness.

Norriss, D.W. & Wilson, H.J. 1986. Greenland White-fronted Geese in Ireland 1985/86. Forest and Wildlife Service, Dublin.

Owen, M., Black, J.M., Agger, M.K. & Campbell C.R.G. 1987. The use of the Solway Firth, Britain by Barnacle Geese Branta leucopsis (Bechst) in Relation to Refuge Establishment and Increases in Numbers. Biol. Conserv. 39: 63-81.

Ruttledge, R.F. & Ogilvie, M.A. 1979. The Past and Current Status of the Greenland White- fronted Goose in Ireland and Britain. Jrish Birds 1: 293-363.

Ryan, J.B. & Cross, J.R. 1984. The Conser- vation of Peatlands in Ireland. Proc. Int. Peat Congr., Dublin 1: 388-406.

Stroud, D.A. 1984a. Status of Greenland White- fronted Geese in Britain 1982/83. Bird Study 31: 111-117.

Stroud, D.A. 1984b, 1985, 1986. Greenland White-fronted Geese in Britain: 1983/84, 1984/85, 1985/86. GWGS, Aberystwyth.

Stroud, D.A., Reed, T.M., Pienkowski, M.W. & Lindsay, R.A. 1987. Birds, Bogs and Forestry: the peatlands of Caithness and Sutherland. Focus on Conservation Series. NCC, Peterborough.

S. Laybourne, Old Schoolhouse, Harpsdale, Halkirk, Caithness KW12 6UN. A.D. Fox, Greenland White-fronted Goose Study, School of Biological Sciences, University College of Wales, Aberystwyth, Dyfed SY23 3DA.

(Present address: The Wildfowl Trust, Slimbridge, Gloucester GL2 7BT).

(Revised ms. received 13 January 1988)

36 = Scottish Birds (1988) 15: 36-39

Mute Swans in Ayrshire

IAIN H. LEACH

The results of a study on Mute Swans in Ayrshire in 1984-86 are presented and compared with the results of censuses in 1955 and 1983. The total population in the county has declined by 59% since 1955. Possible reasons

for this decline are discussed.

Introduction

Concern was expressed during the 1970s about declines in Mute Swan Cygnus olor populations in certain areas of England, with lead poisoning at least partly implicated as a cause of this decline (NCC 1981). Sub- jective impressions suggested that Mute Swans in Ayrshire were also declining at this time. In 1978 the BTO organised a national census based on 10 km squares selected at random (Ogilvie 1981) but, because of the nature of the census, accurate interpretation of the results for Ayrshire was difficult. Continuing concern led to a further com- plete national census in 1983 (Brown & Brown 1985, Ogilvie 1986). The results con- firmed that the population in Ayrshire had declined since the last complete census in 1955 and that breeding success was very poor. This prompted an attempt to survey Mute Swans in detail in Ayrshire in 1984-86 and the results of this study are presented here.

Methods

The study area comprised the old county of Ayrshire. During the years 1984-86 all known breeding sites were visited regularly throughout the year. In addition, all other apparently suitable sites were visited several times annually. As some pairs move their young from the nesting site soon after hat- ching, regular visits allowed these movements to be monitored. To avoid un- due disturbance, no attempt was made to determine clutch size but visits were timed

to count the number of young which hat- ched and regular visits were made subse- quently to assess the survival of cygnets. Counts of non-breeding flocks were made in April when this population is at its most sedentary (Ogilvie 1981).

Statistical results indicated by superior figures are given in the Appendix.

Results

Table 1 summarises the results for 1955 and 1983-86. The number of territorial pairs during 1983-86 remained fairly stable, averaging 24. Although several traditional sites were unoccupied in 1985, five sub-adult pairs held new territories in 1986. Over the four-year period, an average of 61.5% of breeding pairs hatched young. Of 25 breeding attempts which failed during in- cubation, there was strong circumstantial evidence that egg collecting accounted for 15 (60%) of these failures. Of the re- mainder, one nest was flooded, six failed to hatch after prolonged incubation and were presumed infertile whilst three failed for unknown reasons. After 22 failures of first clutches, 10 second attempts were made of which seven hatched young. One pair in 1986 had two clutches robbed but reared four young from a third attempt. The number of young hatched per breeding pair averaged 2.3 over 1983-86 with 1.6 young fledged per breeding pair. Overall cygnet survival was 71.4%, with death due to natural causes thought to account for most of the cygnets lost.

ot

1988

Mute Swans in Ayrshire 37

TABLE 1 Summary of Mute Swan breeding data for Ayrshire, 1955 and 1983-86.

Average

1955 1983* 1984 1985 1986 1983-86 No. of territorial pairs 32 ZS 26 21 25 24 No. of breeding pairs 29 18 22 20 19 20 No. of pairs hatching young 22 10 16 13 10 2. Breeding pairs (%) hatching young 75.9 55.5 TDA. 65.0 52.6 61.5 No. of young hatched 103 23 70 49 40 45 No. of young hatched/breeding pair 3.6 las) Sy Des 2.1 DES No. of young fledged — 11 52 37 30 32 No. of young fledged/breeding pair — 0.6 2.4 1.9 1.6 1.6 Survival (%) of hatched young _ 47.8 74.3 75:5 75.0 71.4

* The figures for 1983 have been adjusted slightly from those of the national census in light of

data obtained subsequently.

Table 2 summarises the numbers of ter- ritorial and non-breeding birds present in 1955 and 1983-86. The non-breeding population averaged 42% of the total population in 1983-86 compared with 70% in 1955.

The ratio of non-breeders to territorial birds in 1955 was very significantly greater! in 1955 than that in any year of the survey. Differences between years during the study were small, although the ratio of non- breeders in 1984 was significantly lower than that in 1985. The only large flock oc- curred on the River Ayr with the remainder

scattered at other, mainly coastal sites. The highest April count on the River Ayr dur- ing 1983-86 was 13 in 1986 whereas in 1955 the count was 68.

Discussion

The 1955 national census (Rawcliffe 1958) provides the only baseline against which comparisons may be made. The total Ayr- shire population was estimated then at 213, with 64 territorial birds and 149 non- breeders. The counts of non-breeding birds were made during May and June and may therefore have been overestimates. In

TABLE 2 Mute Swan population in Ayrshire, 1955 and 1983-86.

Status 1955 Territorial population 64 Non-breeding population 149 Non-breeders as % of total 70 TOTAL POPULATION 213

Average

1983 1984 1985 1986 1983-86 47 53 43 51 49 38 26 48 45 39 45 33 53 47 = 85 79 91 96 88

Notes: Territorial population includes a single territorial male for 1983-86.

38 I.H. Leach

SB 15 (1)

1983-86 the total population averaged 88, revealing an overall decline of 58.7% since 1955. This was mainly due to a 74% drop in the number of non-breeders. The average territorial population over 1983-86 decreas- ed by 23%. The increase in the numbers of non-breeders in 1985 and 1986 may have been partly due to improved breeding suc- cess in the preceding seasons.

The 1978 Mute Swan census showed a decline in the national population of 15% since the last complete census in 1955 (Ogilvie 1981), but by 1983 the population had recovered to a level only 8% below that of 1955 (Ogilvie 1986). In Scotland in 1983 swan numbers were only 3.6% lower than in 1955 (Brown & Brown 1985), but there was considerable variation between regions. Increases of up to 111.1% were recorded in Orkney and the north-east but there were declines of up to 94% in other counties.

It is difficult to find any single explana- tion for the decline in Mute Swans in Ayr- shire. Former breeding sites which are no longer occupied still appear suitable and un- polluted, and few appear to be subject to excessive human disturbance. Overall breeding success during 1983-86 was poor, with only 1.6 young fledged per breeding pair. This compares with 2.8 in Lothian (Brown & Brown 1984), 2.2 in Oxford (Bacon 1980) and 1.5 in the Uists (Spray 1981). As it is mainly locally bred cygnets which maintain both the non-breeding population and, subsequently, the breeding population, it seems likely that poor breeding success may be a major factor preventing an increase in numbers. The most consistently successful pairs used either inaccessible nest sites or sites where human interference was minimal. It is disheartening the egg collecting is still a major problem in the county, particularly as losses at this stage are more damaging than losses of hat- ched young.

Poor weather during the breeding season may reduce breeding performance. In the cold, wet spring of 1983, only 55.5% of breeding pairs hatched young, with 0.6 young fledged per breeding pair. In the

warm and dry spring of 1984, 72.7% of breeding pairs hatched young, with 2.4 young fledged per breeding pair. The highest mortality of young occurred in the first few weeks after hatching (cf. Eltr- ingham 1966). However, the 50% overall mortality quoted by Eltringham is con- siderably higher than the 28.6% recorded in this study.

Although lead poisoning is an impor- tant cause of Mute Swan deaths in certain parts of Britain, Brown & Brown (1984) found no evidence that it was a major pro- blem in Lothian. Insufficient data are available to allow comment on the situation in Ayrshire, but it is likely to be similar to Lothian. Coarse angling is not as popular here as in many parts of England.

Data on Mute Swan mortality are dif- ficult to obtain; the majority of cygnets lost were thought to have died of natural causes, but in 1984 three were killed by a car ferry at Largs and two were shot near Irvine. Analysis of ringing data has shown that about 50% of Mute Swan deaths are due to collisions with overhead power lines (Ogilvie 1967). In Ayrshire at least seven swans are known to have died in this way in 1984. Other factors such as starvation, hard winters and predation by Fox Vulpes vulpes and Mink Lutreola lutreola may also be involved.

The decline in the Ayrshire population since 1955 is probably due to a combination of factors, of which human interference is the most important. Because British Mute Swans are largely sedentary, recovery must come mainly from the local production of young rather than through immigration from unaffected areas. Brown & Brown (1984) felt it unlikely that the Lothian breeding population would return to the levels of the 1950s in the forseeable future. With a relatively small breeding population it seems likely that the same conclusion is applicable to Ayrshire.

Acknowledgments

I should like to thank A.W. Brown and M.A. Ogilvie for supplying the original data from the

1988

Mute Swans in Ayrshire 39

surveys in 1955 and 1983. I also wish to thank the many observers who supplied casual records and R.H. Hogg for supplying the records sub- mitted to the Ayrshire Bird Report. A.W. Brown and G.S. Riddle gave constructive criticism of an earlier draft.

References

Bacon, P.J. 1980. Status and dynamics of a Mute Swan population near Oxford between 1976 and 1978. Wildfowl 31: 37-S0.

Brown, A.W. and Brown, L.M. 1984. The status of the Mute Swan in the Lothians. Scott. Birds 13: 8-15.

Brown, A.W. and Brown, L.M. 1985. The Scot- tish Mute Swan Census 1983. Scott. Birds 13: 1430-148.

Eltringham, S.K. 1966. The survival of Mute Swan cygnets. Bird Study 13: 204-07. NCC. 1981. Report of the Nature Conservancy Council’s Working Group on Lead Poison-

ing in Swans. London.

Ogilvie, M.A. 1967. Population changes and mortality of the Mute Swan in Britain. Wildfowl Trust Ann. Rep. 18: 64-73.

Ogilvie, M.A. 1981. The Mute Swan in Britain, 1978. Bird Study 28: 87-106.

Ogilvie, M.A. 1986. The Mute Swan Cygnus olor in Britain 1983. Bird Study 33: 121-37. Rawcliffe, C.P. 1958. The Scottish Mute Swan Census 1955-56. Bird Study 5: 45-55. Spray, C.J. 1981. An isolated population of Cygnus olor in Scotland. Proc. 2nd Inter- national Swan Symposium, Sappiro, Japan,

Feb. 1980, 191-208.

APPENDIX. Results of statistical tests.

1. 1983, X?=15.5, P<0.001; 1984, X?=31.4, P<0.001; 1985, X*=7.54, P<0.01; 1986, X*=14.11, P<0.001.

2. X?=5.99, P<0.02.

Tain H. Leach, 26 Ashchurch Drive, Wollaton, Nottingham NG8 2RA.

(Revised ms. received 28 January 1988)

40 Scottish Birds (1988) 15: 40-43

The distribution and abundance of the Barn Owl Tyto alba

in south-west Scotland

I.R. TAYLOR, A. DOWELL, T. IRVING,

I.K. LANGFORD AND G. SHAW

Breeding Barn Owls were recorded in 88% of 83 10 km squares examined in south-west Scotland between 1981 and 1986. The total breeding population for the area was estimated to be 139 pairs in 1985 (a poor vole year) and

321 pairs in 1981 (a good vole year).

Introduction

Throughout most of Britain and much of continental Europe Barn Owl Tyto alba numbers have declined considerably in re- cent years (e.g. Bunn ef al. 1982). The decrease has been attributed mainly to changes in farming practice affecting the availability both of suitable hunting areas and of nest sites. In Scotland Barn Owls have been recorded breeding as far north as Sutherland but the main concentration is in the south-west (Thom 1985). The BTO Breeding Atlas (Sharrock 1976) recorded breeding in most of the 10 km squares of this area but did not estimate the density or number of pairs. This paper provides up- to-date information on the distribution of the Barn Owl in south-west Scotland and estimates density and population size.

Methods

Between 1981 and 1986, 83 10 km squares in south-west Scotland were examined. The time required to search each, especially those where Barn Owl numbers were low, was considerable and it was not possible to visit all squares every year. Each was visited at least once during the study and most were visited in at least four of the six years. Evidence of breeding was accepted only when one of the observers had seen a Barn Owl clutch or brood. No use was made of any unsubstantiated information.

Barn Owl numbers are known to fluc- tuate in response to changes in the abun-

dance of their prey (Cramp 1985). To in- clude such variation in the estimates of den- sity, the abundance of the owls’ main prey, the Short-tailed Vole Microtus agrestis was monitored each year by trapping in early spring (March/April) when the owls started to lay. Voles were trapped at six sites each year in young conifer plantations. At each site 25 trapping stations were used with two treadle-operated breakback traps at each set for five consecutive nights. The cumulative catch averaged over the trapping sites was used as the year index of vole abundance. Using the same method Charles (1981) has shown that changes in vole abundance occur synchronously over south-west Scotland.

Owl densities were then examined in the years of highest and lowest vole numbers recorded during the study. Estimates of den- sity were obtained from a sample of 22 10 km squares selected throughout the area from New Galloway to Newcastleton, which included all habitats and altitudes in the area. The sample represented 36% of the total available land area. The number of Barn Owl pairs breeding in each square was counted in 1981, 1983 and 1985.

The method of locating breeding pairs involved searching all man-made structures such as farm buildings and disused and derelict buildings. It is unlikely that pairs nesting in such sites were overlooked. Pairs nesting in trees were more difficult to find. Extensive searches were made in winter to

1988

Barn Owl in SW Scotland 41

oo @@6

\\

RS

10@@0@0e000080 e @

oo ©@ 3 © @ee@52eee6 6 @ @ @ &

—

<= x

gs

oe ee ° Si Tl aan @ Positive 0 30 km Gy Negative

FIGURE 1 The distribution of breeding Barn Owls in south-west Scotland, 1981-1986.

locate suitable nest trees and in summer for occupied sites, concentrating in particular on isolated trees and those along hedges and woodland edges. Sites used by breeding Barn Owls were conspicuous, even in winter, from pellet debris at the base of the tree.

Results

Breeding Barn Owls were recorded in 88% of the 83 squares examined (Fig. 1). Of the 10 squares in which none was found, eight were in mainly upland areas where most of the land was above 1000’ (305 m) and so beyond the normal altitudinal limit of the species. The remaining two were on the coast. One of these, south of Portpatrick on the west of the area, included stretches of sea cliff which could not be searched ex- haustively and nesting pairs may have been missed there. The other was a narrow strip on the north Solway shore west of Annan.

The mean density of breeding Barn Owls in 1981, the year of highest vole abun- dance (index = 44), was 5.1 pairs per 10 km square. In 1985, the year with the lowest vole abundance (index = 1), the mean den- sity of owls was 2.2 pairs per 10 km square (Table 1, Fig. 2). However, high confidence intervals were associated with these estimates due to the great variation in the number of pairs (0-10) in the squares sampl- ed. We can be 95% certain that the true mean density lay between 3.72 and 6.48

TABLE 1 Estimates of the mean density (pairs per 10 km2) and total number of breeding pairs of Barn Owls in south-west Scotland. 95% con- fidence limits are shown in parentheses.

1981 1983 1985 Density 5.1 (1.38) 2.5 (0.82) 2.2 (0.65) No. pairs 321 (87) 156 (52) 139 (41)

42 I.R. Taylor et al.

— > WM -"Ws + * U1 CS <3 > © 4 ©

Index of vole abundance

0

1978 80 82

SB 15 (1)

84 86

FIGURE 2 Yearly changes in the abundance of the Short-tailed Vole.

breeding pairs of owls per 10 km square in 1981 and between 1.55 and 2.88 pairs in 1985.

It is possible to estimate the size of the south-west Scotland breeding population from these values. The total land surface area was approximately 6300 km2. Thus the estimated mean number in 1981 was 321 pairs and in 1985 was 139 pairs (Table 1).

Discussion

This study has shown that the Barn Owl still breeds in most of south-west Scotland ex- cept at higher altitudes.

Because of changes in the Barn Owl population and the percentage breeding, the estimated density of breeding pairs varied more than twofold over a four-year period associated with changes in the abundance of Short-tailed Voles. A detailed analysis of Barn Owl population dynamics will be published elsewhere but it is important to

state here that no factor other than changes in vole abundance could have accounted for these large short-term changes in the owl population. There were no significant changes in land use during the period. Some nest sites were lost but in 1985 when owl numbers were lowest, available nest sites considerably exceeded the number of breeding pairs. The winter of 1981-82 was quite severe but the slightly higher than average mortality associated with it was not enough to account for the 1985 population level. The variation stresses the need to take prey abundance into account when attemp- ting to assess long-term changes in Barn Owl populations.

Also, even when considering only a single year and with a very large sample size (36% of the total area) the confidence in- tervals associated with the density estimates were large (+/— 27% of the mean). This means that even if the prey supply was con- stant it would be impossible to be sure the

et

7

1988

Barn Owl in SW Scotland 43

population had definitely increased or decreased unless the change in either direc- tion exceeded 27% of the previous year’s estimate.

Acknowledgments

We would like to thank I. Leach, Dr M. Mar- quiss, G. Sheppard, R.T. Smith and J.F. Young for providing records of breeding Barn Owls. We are grateful to all the landowners and their agents for access to their land and buildings during the study, and to the NCC, SOC and WWFE for finan- cial assistance.

References

Bunn, D.S., Warburton, A.B. & Wilson, R.D.S. 1982. The Barn Owl. Poyser, Calton. Charles, W.N. 1981. The abundance of the field vole Microtus agrestis in conifer plantations. In: Last, F.T. & A.S. Gardiner (Eds). Forest

and woodland ecology. ITE, Cambridge.

Cramp, S. 1985 (Ed). Birds of the Western Pale- arctic. Vol IV. Oxford University Press, Oxford.

Sharrock, J.T.R. 1976 (Ed). The Atlas of Breeding Birds in Britain and Ireland. Poyser, Berkamsted.

Thom, V.M. 1985. Birds in Scotland. Poyser, Calton.

I.R. Taylor and I.K. Langford, Department of Forestry and Natural Resources, University of Edinburgh, Mayfield Road, Edinburgh EH9 3JU.

A. Dowell, 12 Glentrool Village, Newton Stewart, Wigtonshire DG8 6TB.

T. Irving, The Bield, Wauchope Place, Langholm, Dumfries-shire DG13 OAZ. G. Shaw, Kirriereoch, Bargrennan, Newton Stewart, Wigtonshire DG8 6TB.

(Revised ms. received 14 January 1988)

° 20° is

Sip) e Whines soe ll an

cane’ eS ited Barn Owl

7 AAA

av

Andrew Dowell

44 Short Notes

SB 15 (1)

Short Notes

Sexing and distinguishing juvenile Dotterel in summer

Juvenile Dotterel Charadrius morinellus fall into two categories of size and colour: These are clear among individuals in each brood, as early as when they still retain traces of down on the head and nape. Young in the smaller size category are brownish-grey, with off-white eye-stripes and breast stripes. A good field character is that young in the bigger category are warmer brown, their dark and light parts both being suffused with a creamy, fawn, or cinnamon colouring. They also have a pale, fawn-coloured eye- stripe which is broader relative to body size. The pale colouring in their eye-stripe nar- rows immediately in front of the eye and then broadens into a big patch of similar- coloured pale feathers on the cheek and chin; both cheek and chin usually have fewer greyish feathers than in the smaller- category young. I suggest that the smaller- category young are cocks and the larger ones hens; this could be tested by ringing chicks and noting their sex as adults in later years. The size difference is not surprising, as adult hen Dotterel greatly surpass cocks in size.

Pintail breeding on Colonsay

On 14 May 1986 while visiting Colonsay, Roy and Joan Ramage of Alloa found a pair of Pintail Anas acuta with seven small young, one of which appeared very weak. On 17 May they saw the female with six young. Two pairs of Shelduck Tadorna tadorna, two pairs of Eider Somateria mollissima and a pair of Mallard Anas platyrhynchos also had broods in the same area although of these, only one pair of Shelduck raised young to fledging. There is regular disturbance by humans in the area and adult birds frequently leave their young exposed to predators such as gulls, Larus spp., Crows Corvus corone and Herons Ardea cinerea all of which are plentiful. The Pintails may have had some advantage over the other ducks by their more

Other species where adult cocks and hens differ greatly in size show big sexual dif- ferences in the body size of feathered young in each brood; these are easily proven in some cases such as Capercaillie Tetrao urogallus by the fact that young cocks grow blackish feathers long before the brood is fully fledged.

BWP Vol III states ‘‘juvenile separable at close range’’, but in summer it is easy to tell juvenile from adult at over 100 m. Wat- son (SB 4: 179-203) wrote ‘‘fully grown young are remarkably unlike the adults in plumage, having a generally creamy ground colour with heavy blackish marks on the wings and back’’. The predominant colour on their wings and back is much darker, and most of the feathers there are edged narrow- ly with a paler colouring than in adults, giv- ing a far more contrasty, dappled pattern. The difference is so great that juveniles at first sight might seem a different species.

Adam Watson, Institute of Terrestrial Ecology, Banchory AB3 4BY

nocturnal habits (BWP Vol I). Although the female was seen on several occasions, the young were more difficult to find. Sightings of the female with six young on 30 May (J. Clarke) and four large young on 4 July (G. Riddle) were reported. The male was seen on only one other occasion, on 14 July (D. Gib- by). This is the first confirmed breeding record for Pintail on Colonsay. There have been only five previous recorded sightings for the island. At most c.20 pairs breed in Scotland; breeding is sporadic except in Orkney and Caithness (Thom V.M. 1986. Birds in Scotland. Poyser, Calton).

J. & P.M. Clarke, Isle of Colonsay, Argyll

1988

Short Notes 45

Mead & Pepler (BB 68: 89-99) documented predation of Sand Martins Riparia riparia at breeding colonies during the BTO ringing enquiry into this species. Black-headed Gulls

_ Larus ridibundus were recorded approaching

_ martin nestholes at several colonies, although

the gulls were thought to be feeding on fleas. Black-headed Gulls were also noted taking the contents of exposed nests after cliff col- lapses, but there was no evidence of gulls as active predators rather than scavengers at

- Sand Martin colonies.

- During 1982 a colony of about 37 pairs of Black-headed Gulls nested beside a colony of c.920 pairs of Sand Martins at Barbush

_ sand quarry, near Dunblane in Central Scot-

land. The gulls were the most frequently observed predators of martins at the colony that year, and at least five chicks were taken. Adult gulls typically approached burrow entrances where full-grown Sand Martin

Two cases of a yearling and an adult

On 31 May 1986 we found a Merlin Fa/co columbarius nest with four eggs on the ground on a grouse moor in lower Deeside, Grampian. A yearling male (i.e. in brown plumage) called the female off the nest and was seen passing food to her. He then went to the nest. About five minutes later, after eating the prey, the female returned to the nest and the male left. When we examined the nest 30 minutes later the female and the yearling male both called in alarm. On 21 June a pair alarm-called but the male was notaged. On 7 July only the female was pre- sent, sporadically, during our four-hour Visit to the area. On 16 July four young had fledged and a full adult male (i.e. in blue- grey plumage) delivered food to one of them. This male and the female were pre- sent for the next hour.

Similar behaviour was observed in 1987 at a young conifer plantation 30 km away. A pair of Merlins display-called over the ter- ritory (T1) on 26 April. During 1 and 2 May the male at Tl was seen to be a yearling. On 26 April at a neighbouring territory 1

_ Black-headed Gulls as predators of breeding Sand Martins

nestlings awaited feeding, and if successful, pulled the nestlings from the burrow and carried them away. The martin nestlings nor- mally retreated from their burrow entrance if a gull appeared, and gulls were mobbed by adult martins. Most of the gulls’ attempts at taking nestlings were unsuccessful. Only one nestling was taken at each successful predation attempt. One nestling aged about 20 days was forced to be dropped by a fly- ing guli, but was dead on retrieval.

Carrion Crows Corvus corone employed similar tactics to the gulls at Sand Martin burrows, although only the gulls, and on one occasion a Kestrel Falco tinnunculus were noted as successful predators.

Gareth Jones, Department of Zoology, University of Bristol, Woodland Road, Bristol BS8 1UG

male Merlin attending the same nest site

km away (T2) a pair was established and the male was in full adult plumage. He pro- spected an old Crow Corvus corone nest on 1 May and ‘scraped’ it on 2 May. A female was present both times. On 8 and 16 May no Merlins were seen or heard at T2 but in Tl on 16 May a yearling male was in- cubating five eggs in an old Crow nest in a Scots pine Pinus sylvestris. When disturb- ed the male was soon joined by a female and both called. On 7 June T2 was unoccupied and T1 still active with the yearling male and the femaie defending it. On 16 June all the eggs had hatched and the young were about one week old. The female was very ag- gressive during this five minute visit; no male was present. On 29 June a male in full adult plumage flew from the nest tree call- ing. The female arrived with prey three minutes later and both then alarm-called. They were present for the next 40 minutes, while the five young were weighed, measured and ringed. On 8 July during a three-hour watch the adult male (twice) and the female delivered food to the five recently

46 Short Notes

SB 15 (1)

fledged young. The first delivery from the male involved him calling to and being answered by the young for fully three minutes. On 12 July four young were seen flying strongly and the adult male brought food to one of them. For all the males (ex- cept on 21 June 1986) we had excellent close range views using good quality optics. Yearling male Merlins begin a partial moult during February to May and subse- quently complete their moult into adult plumage between September and November (Moult in Birds. BTO Guide 19). At other nests in Grampian where the male was a yearling, each was still in brown plumage at the end of the breeding season in late Ju- ly. In the closely related and similarly sized Kestrel F. tinnunculus yearlings do not com- plete their moult into adult plumage until October/November of their second year, and raptors in general have a long moult period (BTO Guide 19). It is therefore ex- tremely unlikely that the yearling males at the nests we have described could have moulted into full adult plumage between our observations. Male Merlins have twice been reported as being replaced by another male during the same breeding season (Newton, I. 1979. Population Ecology of Raptors. Poyser, Berkhamsted). This happened after the original males had been shot. We have no evidence for persecution at the Grampian

Sparrowhawk specialising on Swifts Every year a pair of Sparrowhawks Accipter nisus nests on an island on the River Don in Seaton Park, Aberdeen. This site was one of eight I visited in 1987 to collect prey remains for analysis.

On | August 1987 these included Grey Wagtail Moticilla cinerea, Chaffinch Fringilla coelebs, Wren Troglodytes troglodytes and Blackbird Turdus merula, but it was surprising also to find the remains of 12 Swifts Apus apus. Two were in the nest and consisted of wing bones with outer primaries attached. The 10 found on the

nests, and it is very unlikely. In 1986 the oc- cupation of the eight surrounding territories seemed to rule out any surplus adult males locally; four held pairs (three with adult males and one with a yearling), two held single females and two were unoccupied. The food delivery on 16 July 1986 could have been instinctive, perhaps by a neighbouring male or a failed breeder from further afield. After prospecting the Crow nest on 1 May 1987 the adult male flew to Tl and displayed there. We believe he eventually moved from T2 to T1 either after failing to breed or an early breeding failure. There were probably no other surplus adult males nearby as three of the five surroun- ding territories were unoccupied.

Unfortunately our observations were too few to decide whether the yearling males died, left voluntarily, were displaced, or briefly, were acting as ‘helpers’ (although in each case the younger male initiated the nest). It would be interesting if other Merlin enthusiasts could monitor extensively nests with a yearling male in attendance.

We should like to thank Eric Meek and the editorial panel for their comments on this note.

G.W. Rebecca*, B.L. Cosnette and A. Duncan

*31 Rainnieshill Gardens, Newmachar, Aberdeenshire ABS O0JZ.

ground each consisted of wing feathers and a few breast feathers.

I. Newton (1986, The Sparrowhawk, Poyser, Calton) found only one Swift in his analysis of 10,000 kills. It seems clear that the Aberdeen bird specialised in catching Swifts. The nest site overlooked rapids where Swifts were often seen flying low over the water. Individual Swifts were also found at two other Sparrowhawk sites in Aberdeen.

Judy M. Cooper, 14 Abbotshall Gardens, Cults, Aberdeen ABI 9LA

.————--—_— ———————_-- ——— ——— oo —iannaeninnne re

| 1988 Items of Scottish Interest 47

Items of Scottish Interest

) Articles and Reports on birds in Scotland, mainly on status and distribution. (Previous lists | 14(3) 189-190; 14(4) 217-218). References from the widely available journals British Birds, _ Bird Study and Ringing and Migration are excluded. Most of these items are available for | reference in the Waterston Library. Items marked with an asterisk are also available from | the SOC Bird Bookshop postfree to SOC members at the prices quoted.

_ The librarian is glad to receive reprints or copies of papers on any aspect of ornithology

_ or general natural history.

| Scientific papers.

_ Changes in numbers of cliff-nesting seabirds in Orkney, 1976-1985. S. Benn, M.L. Tasker & A. Webb 1987. Seabird 10: 51-57.

A Survey of moorland birds on the Campsie Fells/Touch Hills massif, Stirling in 1987. J. Calladine, S. Dougill, K.B. Shepherd, D.A. Stroud, J. Turner & C.M. Crawford 1987. Nature Conservancy Council, Nor- thminster House, Peterborough (Contract Report) 58 pp.

The Outer Hebrides. P. Cunningham 1987. Sea Swallow 36: 20-24.

Opportunistic feeding of Black Guillemots at fishing vessels. P.J. Ewins 1987. Seabird 10: 58-59.

Sandeels in the diet of the Fulmar in Shetland, Scotland. J.A. Fowler & A.P. Dye 1987. Seabird 10: 71-74.

The marine distribution of Sooty Shearwater, Manx Shearwater, Storm Petrel and Leach’s Petrel in the North Sea. A.J. Hall, M.L. Tasker & A. Webb 1987. Seabird 10: 60-70.

Chick production at Kittiwake colonies in Shet- land, 1986. M. Heubeck, P.V. Harvey & I.S. Robertson 1987. Seabird 10: 34-42.

Birds of Cape Wrath. D. Law 1987. Adjutant 17: 10-14.

Aggression in shorebirds in relation to flock density and composition. N.B. Metcalfe & R.W. Furness 1987. Ibis 129: 553-563. A study of wintering waders in the Firth of Clyde covering three winters.

Breeding performance of Merlins in North-East Scotland in 1986. G.W. Rebecca 1987. Hawk Trust Annual Report 1985-86 15: 12-13.

Colonization of island woodlands in the Heb- rides of Scotland. T.M. Reed 1987. Acta Oecol. 8: 275-280.

Effects of aerial applications of Fenitrothion on bird populations of a Scottish pine plan- tation. C.J. Spray, H.Q.P. Crick & A.D.M. Hart 1987. J. Appl. Ecol. 24: 29-47.

Social structuring at a communal roost of Choughs. E. Still, P. Monaghan & E. Bignal 1987. Ibis 129: 398-403.

Census of breeding Wigeon in Ettrick Forest. D.L. Thomson 1987. Nature Conservancy Council, 12 Hope Terrace, Edinburgh 54 pp.

Numbers, territory size and turnover of Short- eared Owls in relation to vole abundance. A. Village 1987. Orn. Scand. 18: 198-204. A study at Eskdalemuir, southern Scotland.

Weight and sexual cycle of Scottish Rock Ptarmigan. A. Watson 1987. Orn. Scand. 18: 231-232.

The use of models to influence the grazing sites chosen by Barnacle Geese on Islay, Scotland. X. Zhu, D.C. Houston & S. Per- cival 1987. Wildfowl 38: 46-48.

The Royal Air Force Ornithological Society Expedition to Berneray and Mingulay, June- July 1985. N.A. Smith, B.R. Withers & K.W. Earnshaw 1987. Royal Air Force Orni- thological Society Journal 17: 1-83.

Bird Reports

Argyll Bird Report for 1986. (78 pp) Argyll Bird Club, C.A. Galbraith & A.R. Jennings (Eds) 1987. *£3.50. This is the 4th report of the Argyll Bird Club. It includes a 37-page species list, a ringing report and several short articles on Corncrakes, Mink, Golden Eagles and on various aspects of forestry and birds.

Borders Bird Report for 1986. (56 pp) R.D. Murray (Ed) 1987. *£2.50.

48 Items of Scottish Interest

SB TS%1)

Fife and Kinross Bird Report for 1986. (34 pp) Fife Bird Club. 1987. *£2.50. Includes a species list, and short reports on seawatching at Fife Ness, etc.

Flannan Islands visits. A.D.K. Ramsay. Unpub- lished report 3pp.

Lanarkshire Bird Report for 1986. (21 pp) I. English (Ed). An unpublished report for the RSPB Hamilton Area Members Group.

Lothian Bird Report for 1986. (84pp) M.R. Leven & I.J. Andrews (Eds) 1987. *£3.50. The 50-page systematic list includes many useful tables of peak counts of birds, an 8-page report on a census of breeding Little Grebe, Great Crested Grebe and Tufted Duck in the Lothians, and short reports on breeding Fulmars on the East Lothian mainland, on breeding Mute Swans, Goose counts and Water Rails.

North-East Scotland Bird Report for 1986. (52 pp) M.L. Tasker (Ed) 1987. *£2. Includes a detailed species list, and a 9-page report on seabirds breeding on the coast of North-East Scotland.

Stewartry and Wigtown Districts Bird Report for 1986. (17 pp) A.D. Watson (Ed) 1987. *£1.95. The 2nd annual report for this part of Dumfries and Galloway Region.

Tay Ringing Group Report for 1984-86. (64 pp) Tay Ringing Group, Dundee 1987. 43 pages of this report are studies of Dunnocks, Sedge Warblers, Sandmartins, Siskins and Mute Swans.

Dungavel Area Bird Report for 1983-1986. (43 pp) R. Morton & I. Livingstone (1987). A detailed study of an area of moorland on the borders of Lanarkshire and Ayrshire which has been increasingly afforested over the period. Unpublished report.

Holyrood Bird Records 1986. (5 pp) L.L.J. Vick 1987. A brief account of birds seen in Holyrood Park, Edinburgh, which includes Duddingston Loch and the Bawsinch Reserve. Unpublished report.

W.G. Harper

Merlins

B.L. Cosnette

Advice to Contributors

_ | Authors should bear in mind that only a small : gieerortion of the Scottish Birds readership is

science-trained, and should aim to present their | material concisely, interestingly and clearly. | Unfamiliar technical terms and symbols should _be avoided wherever possible and if deemed essen- tial should be explained. Supporting statistics _should be kept to a minimum. All papers and short notes are accepted on the understanding that | they have not been offered for publication elsewhere and that they will be subject to editing. | Papers will be acknowledged on receipt and will _be reviewed by at least two members of the / editorial panel, and in some cases also by an in- | dependent referee, before being accepted. They will normally be published in order of acceptance | of fully revised manuscripts. The editors will be happy to advise authors on the preparation of | papers. | Reference should be made to recent issues | of Scottish Birds for guidance on style of presen- tation, use of capitals, form of references, etc. Papers should be typed on one side of the paper | only, double-spaced and with wide margins; two | copies are required and the author should also retain one. Headings should NOT be underlined, | nor typed entirely in capitals. Scientific names

should follow the first text reference to each species and should follow Voous’ ‘List of Recent Holarctic Bird Species’ as given in The British Birds’ List of Birds of the Western Palearctic (1984).

Tables, maps and diagrams should be desig- ned to fit either a single column or the full page width. Tables should be self-explanatory and headings should be kept as simple as possible, with footnotes used to provide extra details where necessary. Each table should be on a separate sheet. Maps and diagrams should be in Indian ink and drawn so as to permit reduction to half their Original size. If necessary they may be submitted without lettering and accompanied by a photo- copy showing the lettering required. Captions should be typed on a separate sheet. Relevant line-drawings (in ink) will be welcomed, as will photographs (preferably black & white glossy prints).

Authors are responsible for checking their own proofs and returning them promptly. Text changes (as distinct from correction of printer’s errors) at the proof stage involve extra cost and will only be accommodated under exceptional circumstances.

Scottish Birds

Volume 15 Part 1 June 1988

Contents

The recolonisation of the Isle of May by Common and Arctic Terns.

S. Wanless

The distribution and status of Arctic and Great Skuas in Shetland 1985-86. P.J. Ewins, P.M. Ellis, D.B. Bird & A. Prior

The seabirds of St Kilda, 1987. Mark L. Tasker, Peter R. Moore & Richard A. Schofield

Greenland White-fronted Geese in Caithness. S. Laybourne & A.D. Fox

Mute Swans in Ayrshire. Jain H. Leach

The distribution and abundance of the Barn Owl Tyto alba in south-west Scotland. I.R. Taylor, A. Dowell, T. Irving, 1.K. Langtord & G. Shaw

Short Notes

Sexing and distinguishing juvenile Dotterel in summer. Adam Watson

Pintail breeding on Colonsay. J. & P.M. Clarke

Black-headed Gulls as predators of breeding Sand Martins. Gareth Jones

Two cases of a yearling and an adult male Merlin attending the same nest site. G.W. Rebecca, B.L. Cosnette & A. Duncan

Sparrowhawk specialising on Swifts. Judy M. Cooper Items of Scottish interest. W.G. Harper

Published by the Scottish Ornithologists’ Club, 21 Regent Terrace, Edinburgh EH7 5BT. © 1988

Printed by Alexander Ritchie & Son Ltd., Edinburgh.

2A

30

36

40

- — $e SSae —— a = ego —— Tee an = = S ——— en ==, ; bs : = _ —— .—__— a: Ee Bs at ——— < am x — _—_ a oe Sa - Ra + —_ ii ~ —_— —- oT a |, : <a aa a, Gaia 7 is —_ ae & = = — — ~ a hep W ei W oe ae => =) = = — omen —= = a — Qe? ee) SO Ee _—— wa

SX

WV

Scottish Birds

The Journal of the Scottish Ornithologists’ Club

Editor: N. Picozzi Assisted by: N.P. Ashmole and M.A. Ogilvie. Business Editor: The Secretary, S.O.C., 21 Regent Terrace, Edinburgh EH7 SBT.

Scottish Birds, the official journal of the Scottish Ornithologists’ Club, publishes original material relating to ornithology in Scotland; papers concerned with status and distribution are particularly invited. Short notes are also accepted. Papers and short notes should be sent to The Editor, Scottish Birds, 21 Regent Terrace, Edinburgh EH7 SBT.

Two numbers of Scottish Birds are published each year, in June and December. Scot- tish Birds is issued free to members of the Scottish Ornithologists’ Club, who also receive the quarterly newsletter Scottish Bird News. Scottish Birds and Scottish Bird News are available to non-members at a subscription rate (1988) of £30.00. For information on advertising con- tact the Business Editor.

The Scottish Ornithologists’ Club was formed in 1936 to encourage all aspects of ornithology in Scotland. It has local branches which meet in Aberdeen, Ayr, the Borders, Dumfries, Dundee, Edinburgh, Glasgow, Inverness, New Galloway, St Andrews, Stirling, Stranraer and Thurso, each with its own programme of field meetings and winter lectures. The Waterston Library at the Club’s headquarters at 21 Regent Terrace, Edinburgh EH7 SBT is one of the most comprehensive ornithological libraries in Scotland, and is available for reference dur- ing office hours (Monday to Friday, 9am to Spm). The SOC Bird Bookshop at Regent Ter- race carries a large stock and can supply any Natural History book in print. Much of the business is by mail order, but personal callers are welcomed during office hours. The current catalogue is available on request. The Bird Bookshop is run by the SOC for the benefit of ornithology in Scotland.

SOC annual membership subscription rates

ASGUIE seed ccccscc essence sea ddive es cncnemeepmemenadxcr occas toe daeedaeaueea doce eee £10.00 Junior (under 18, or student under 25) ....ccccccccc...cccccmecmne ce vce ccccseeeeneeee ee emaaan £4.00 Family (husband & wife, & children under 18) ... 2.2. --c2.2e.c.ce- oss enecene seen £14.50 | OMS ICGINSIG WI! sonseocoscecodeoucconee £200.00 PAMULY | coccescicccccstaceesecs ee £300.00

Pensioner Wni@ividual socasscssecceaeeeee £6.00 FAMULY: scones so0ceseeecucienes eee eee £8.50

es

Scottish Birds (1988) 15: 49-60

49

Numbers of wintering Pink-footed and Greylag Geese in north-east Scotland 1950-1986

M.V. BELL, J. DUNBAR AND J. PARKIN

The status of Pink-footed and Greylag Geese in north-

_ east Scotland this century is described with emphasis on _ the large changes which have occurred since 1960. Dur- ing this period the numbers of Pinkfeet have increased greatly at the two traditional roosts, the Loch of

_ Strathbeg and the Meikle Loch. Up to 30% (40,000 birds) of the British wintering population may roost there. Greylag have also increased and now occupy five new roosts. North-east Scotland regularly holds up to 30% (30,000 birds) of the British wintering population of - Greylag. The reasons for these changes are discussed.

Introduction

Pink-footed Geese Anser brachyrhynchos and Greylag Geese Anser anser have been wintering in north-east Scotland for many years. The early Scottish naturalist Thomas Edward recorded Pinkfeet at the Loch of Strathbeg (Smiles 1877), but there was con- siderable confusion between this species and the Bean Goose Anser fabalis in the 19th century. Pinkfeet were certainly present at Strathbeg by the turn of the century and by the late 1930s up to several thousand occur- red here, and in the Slains area to the south in autumn (Berry 1939). However, the use of the Loch of Strathbeg as a roost was pro- bably interrupted during both world wars. In the Great War it was used as a seaplane base and in the second war there was distur- bance from the adjacent Crimond airfield. At the time of Berry’s review of wild duck and wild geese in Scotland, Greylag were stated to be present in only small numbers at several sites in the Dee faunal area, with their status unchanged for many years (Berry 1939). No quantitative information on the wintering geese existed before the Wildfowl Trust began its series of national November counts of Pinkfeet in 1950 and Greylag in 1960.

This paper describes the fluctuations in numbers of both species since then. During this period several new roosts have been oc- cupied by Greylag, and the last 10 years in particular have seen some dramatic changes. North-east Scotland is now one of the most important areas in Britain for both Pinkfeet and Greylag with up to 40,000 and 30,000 respectively.

Survey area

The area of north-east Scotland discussed in this paper is shown in Fig. 1. It comprises the area north of the River Dee and east of the River Deveron, mainly the old county of Aberdeenshire including part of Kincar- dine to the south and Banff to the north- west, all now in Grampian region. This large area has very few stretches of water which are suitable as goose roosts and consequently large numbers of geese are concentrated in a few sites. This has the advantage that a few observers can obtain a fairly complete picture of the changes in numbers through the winter. There are two major Pinkfeet roosts, at Loch of Strathbeg and the Meikle Loch/Ythan estuary, and seven Greylag

50 M.V. Bell et al.

Deveron

Wells of Ythan

Howe of

Howe of Tarland Dinnet e Lochs @

Auchlossan

Ballogie

SB 15 (2)

Loch of ®&®

Strathbeg

Fedderate Reservoir

Meikle Loch

© Hslains Lochs e

Corby/Loch

Loch of Skene

FIGURE 1. A map of north-east Scotland showing the position of the goose roosts and other places mentioned in the text, in relation to the main rivers and the coast. Pinkfeet roost at Loch of Strathbeg and the Meikle Loch, and Greylag roost at the Loch of Strathbeg, Fedderate Reservoir, Haddo House Lake, the Slains Lochs, Corby Loch, Loch of Skene and the Dinnet Lochs.

roosts, at Loch of Strathbeg, Loch of Skene, Lochs Davan and Kinord (collectively the Dinnet Lochs), Haddo House, Corby Loch, the Slains Lochs and Fedderate Reservoir.

The roosts in north-east Scotland are well separated from the other wintering areas for both species. There are no major Pinkfoot sites to the north and the nearest roost to the south is the Montrose Basin (77 km SSW of the Ythan estuary), which is regaining its former importance since the Scottish Wildlife Trust established a reserve there in 1981. There are major concentra- tions of Greylag in Caithness, Easter Ross and along the Moray Firth. The nearest

roosts are at Loch Spynie near Elgin (82 km west of Strathbeg) and in Strathmore (Per- thshire and Angus), 50 km to the south of the Dinnet Lochs.

Methods

Experienced observers familiar with particular sites counted geese arriving at or leaving the roosts. Long runs of data were available from the same few observers for most roosts.

The national November count data were taken from Boyd & Ogilvie (1969, 1972), Ogilvie & Boyd (1976), from the annual reports of the Wildfowl Trust and from Dr M.A. Ogilvie (pers. comm.).

y f ;

EE OE

Geese in NE Scotland 51

Agricultural statistics on cropping regimes were obtained from the Scottish Records Office for selected Aberdeenshire parishes and years since 1960. The parishes of Crimond, Lonmay

_ and St Fergus were taken as representative of the _ feeding areas used by the geese from Strathbeg; those of Slains, Logie-Buchan and Foveran for _ the geese from Meikle; Methlick and Meldrum _ for the geese from Haddo; Kinellar and Skene for

the geese from Skene and Lumphanan, and Leochel-Cushnie for the geese from Dinnet. The years 1960, 1965, 1970, 1975, 1978, 1980, 1982 and 1984 were taken to give a spread over the period since north-east Scotland became of ma- jor importance for wintering geese.

Results

Buchan

Loch of Strathbeg Of the two goose roosts in Buchan, the Loch of Strathbeg qualifies as a site of international importance for ducks, geese and swans. The loch is large (220 ha), shallow, rather exposed and within 500 m of the sea. It is a dif- ficult roost to count because of its size. Since the RSPB reserve was established in 1973 most of the geese have roosted at the sheltered north-west end of the loch in the peninsula area which is relatively undisturbed by wildfowlers.

There is a run of monthly counts at Strathbeg since the mid-1950s. Fig. 2 shows the winter max- imum and average mid-monthly count (October to March up to 1973, October to April since 1973) for both Pinkfeet and Greylag at Loch of Strathbeg since the 1954/55 winter. Table 1 shows the average mid-monthly counts for seven con- venient periods from 1954/55 to 1986/87 and the maximum count within each period. Until the 1963/64 winter numbers of both species were low, with Pinkfeet slightly the more numerous. Greylag numbers were relatively stable through the winter while Pinkfeet numbers reached a peak in autumn and then declined. Numbers remain- ed fairly constant for the next eight or nine winters with both species present in similar numbers and increasingly likely to stay through the winter.

Numbers increased after 1973. In the follow- ing 10 years the Greylag reached maxima of 9500 in November 1975, 9600 in November 1982 and 8350 in March 1985. Since the 1984/85 winter the Greylag numbers have decreased slightly but Pinkfeet numbers have increased markedly. The maximum count of Pinkfeet at Strathbeg up to 1983/84 was°7500 in November 1977 but in the 1984/85 winter, 20,200 were recorded in October,

Pinkfeet (x 1000)

25 20 15 10 5 o LZ

1954/55 60/61 65/66 70/71 75/76 80/81 85/86

10

0 =

1954/55 60/61 65/66 70/71 75/76 80/81 Winter

FIGURE 2. The average monthly count and maximum count of Pinkfeet and Greylag at Loch of Strathbeg each winter from the 1954/55 - 1986/87 winters.

85/86

Greylag (x 1000) oO

with 27,900 in October 1985 and 29,800 in Oc- tober 1986. Over 10,000 were also present in the springs of 1985, 1986 and 1987.

Fedderate Reservoir The small flock of Greylag at Fedderate Reservoir has been counted in November and March since 1977 but it is not clear when the site was first occupied (John Edelsten pers. comm.). This small water (c.10 ha) is easi- ly disturbed and appeared to hold more birds in the spring after the end of the shooting season than in the autumn. The numbers in November ranged from 10-2500 (average 684) between 1977 and 1986, and in March ranged from 1400-2700 birds (average 1840) between 1982 and 1986. The surrounding land which rises to 200 m (650 ft) is very exposed, often taking the brunt of winter snow storms from the north, and is therefore fre- quently deserted by geese in mid-winter.

Ythan Valley

Slains Lochs and Ythan Estuary The Slains estate and adjacent Ythan Estuary have been used ex- tensively by both Pinkfeet and Greylag for many

52 M.V. Bell et al.

SB 15 (2)

TABLE 1. The average monthly counts of geese at Loch of Strathbeg from 1954/55-1986/87, together with the maximum count within each period (no counts available for the 1956/57 winter; nc = not

counted).

PINKFEET Oct Nov Dec 1954/55 - 59/60 1050 670 80 1960/61 - 64/65 1120 1306 180 1965/66 - 69/70 2556 2310 440 1970/71 - 74/75 2868 2870 60 1975/76 - 79/80 4240 5570 3860 1980/81 - 83/84 5725 4288 2975 1984/85 - 86/87 25933 12283 6050

GREYLAG Oct Nov Dec 1954/55 - 59/60 350 398 300 1960/61 - 64/65 876 1432 680 1965/66 - 69/70 1340 2544 770 1970/71 - 74/75 2970 3600 874 1975/76 - 79/80 2986 6950 4160 1980/81 - 83/84 3738 5488 4063 1984/85 - 86/87 1383 3483 3617

years. Unfortunately there is little information on numbers here apart from the national November counts. Since 1980/81 more frequent counts have been carried out.

During the last 10 years the main Pinkfoot roost has been the Meikle Loch (26 ha). This is a Site of Special Scientific Interest (SSSI) but is not part of the Sands of Forvie and Ythan Estuary National Nature Reserve (NNR). The much smaller Cotehill Loch (3.1 ha) and Sand Loch (4.4 ha) are included in the NNR. The Meikle, Cotehill and Sand Lochs are collectively referred to as the Slains Lochs since they are situated on the Slains estate. In the following account Meikle Loch is used when discussing Pinkfeet and the Slains Lochs when discussing Greylag as small numbers of the latter often roosted on Cotehill and Sand Lochs, and pools on the Sands of Forvie. There was controlled wildfowling for geese and ducks at the lochs, and they provided a better sanctuary than the Ythan Estuary NNR 3 km to the south which was regularly disturbed by wildfowling. However, at the end of the shooting season the Pinkfeet resorted increasingly to the estuary and by mid-April most roosted there. Sometimes Pinkfeet roosted on Sand Loch near Collieston, and also on a semi-permanent floodwater pool 1 km to the north of the Meikle Loch.

Jan Feb Mar Apr Maximum 100 0 200 nc 3000 240 305 546 nc 2239

1150 850 2240 nc 4600

1410 1000 1970 nc 5500

2300 2050 2300 3610 7500

1913 3950 2575 4513 7400

2467 4767 8717 10767 29800

Jan Feb Mar Apr Maximum 503 3 52 nc 1100 510 416 ay; nc 3734

1450 1332 1050 nc 4380

1302 1450 2970 nc 5000

2620 2450 1180 1920 9500

1850 3338 3750 3475 9600

2083 5367 4433 683 8350

The much smaller flock of Greylag was also usually found on the Meikle Loch, but showed a greater tendency to disperse. Flocks of several hundred birds quite often occurred on Sand Loch, pools on the Sands of Forvie, and at the Ythan

(e) (e) (@) _ 10 x oO) oO 5 > © _ O 4 SS en 1955 60 65 70 75 80 85 S je) or Xae5 ~ i) Le Sen ~ ¥ = 1955 60 65 70 75 80 85 a Year

FIGURE 3. The November counts of Pinkfeet and Greylag at the Slains Lochs between 1953 and 1986.

; | | | |

Geese in NE Scotland 53 16 14 12 12 10 10 8 Zs be 8 6 6 < 2 2 ® 0 ) we Ss 70 N (a) J F M A M Suvo N (9) J F M AM = 1980/81 a 1981/82 % 18 18 = iG 16 a. te 14 SR 12 = 10 10 8 8 6 6 4 4 2 2 2 Ss 2) N 0 J F M A M p S Oo N 19) J F M A M 1983/84 1984/85 1985/86

FIGURE 4. The numbers of Pinkfeet roosting at the Meikle Loch/Ythan Estuary between September and May from the 1980/81 — 1985/86 winters. The arrows indicate snowfalls and the bars periods with 10 cm or more snow cover. Broken lines are extrapolations based on counts before and after

snowfalls.

Estuary in the spring; smaller numbers roosted at Cotehill less regularly. Site preferences changed between and within winters.

The November counts of Pinkfeet in the 1950s ranged from 0-2000 (Fig. 3), but since the mid-1960s very large numbers have occurred in some autumns. The November counts indicate lit- tle change in the status of the Pinkfoot here over the last 20 years; more than 10,000 birds were pre- sent in November 1966, 1967, 1969, 1974, 1975, 1983 and 1985. It seems probable that more now remain through the winter and occur in the spring than formerly.

Between the 1980/81 and 1985/86 winters detailed counts were obtained of the geese roosting at the Meikle Loch and Ythan Estuary. Whenever possible, attempts were made to count the roost weekly or fortnightly. Small numbers of Pinkfeet often arrived in the second week of September but the main arrival was in late September and early October, usually peaking at c.15,000 birds in October (Fig.4). In most years numbers had dropped by the November count, although high numbers were maintained through November in 1983 and 1985. The main factor in- fluencing numbers in December, January and February was the presence of snow deeper than c.10 cm. In winters when the Buchan coast was snowfree, numbers remained high throughout the winter, e.g. 1982/83, 1983/84 and 1984/85 (Fig. 4). In the three snowy winters, all the Pinkfeet

left at least once, and this is probably more typical. However, the 1985/86 winter was unusual as three well-separated snowfalls caused major departures of the Pinkfeet, but on each occasion they soon returned.

In late February numbers began to build up and the spring peak was reached in late March or April (Fig. 4). In the first three winters of this series (1980/81 — 1982/83) the totals in late February and March were obtained from field counts, which at the time were thought to be ade- quate but which we now think may have been in- complete. No counts were carried out in April 1981 or 1982. In 1984/85 and 1985/86 there were very large peaks in late April and the departure in April 1985 especially was very late. In both years a prolonged period of cold north-westerly winds in late April may have delayed departure.

The numbers of Greylag at the Slains Lochs have been more variable than those of Pinkfeet over the last 25 years. The November counts bet- ween 1960 and 1966, and from 1974 onwards were rather low (326-1783 birds, Fig. 3), but in 1967-1973 numbers were much higher with 9370 in November 1969. The largest numbers of Greylag here since 1974 were 3600 on 9 April 1978 and 3200 on 20 November 1983. The autumn numbers have decreased and the main build-up of Greylag at Slains occurred after the November count in the 1983/84, 1984/85 and 1985/86 winters. Over the six winters studied the average

54 M.V. Bell et al.

SB 15 (2)

number of Greylag present from mid-October to mid-April ranged from 500-1600.

Haddo House Lake The lakes at Haddo House are small (c.4 ha each) and largely surrounded by trees. The Greylag roost on the upper lake which has been in the Grampian Region Country Park since 1981 and a sanctuary for many years. Some duck shooting occurs on the lower lake. The roost was first noticed in the autumn of 1981 when the ranger reported c.4000 Greylag in mid-November. By the end of November c.7000 were present which at the time was one of the largest counts recorded in north-east Scotland.

The early history of this roost is unclear. Ap- parently small numbers (less than 1000) had been present since about 1974. In the winter of 1980/81 they greatly increased. There had never been any shooting of the geese. After the ‘‘discovery’’ of this roost in 1981 it was monitored fairly regularly and numbers fluctuated greatly. In autumn 1982 the Greylag arrived in mid-October, c.4000 were present from late October to mid-November after which numbers increased rapidly to 11,900 on 11 December, falling thereafter. In the 1983/84 winter, numbers showed a similar pattern (peak 7700), but in the 1984/85 winter the autumn peak did not occur and numbers remained at c.4000- 5000 from late October to late January then decreased. In the 1985/86 winter c. 1000 birds were present throughout. Table 2 shows the numbers of Greylag at the Haddo House roost on the November and March counts between the 1981/82 and 1985/86 winters. Pinkfeet were only once recorded at this roost.

TABLE 2. The national November and March counts of Greylag Geese at the Haddo House roost from the 1981/83-1986/87 winter.

1981/82 1982/83 1983/84 1984/85 1985/86

Nov 4000 4650 4600 4200 900 Mar 3500 4000 4500 650 500

The reason for these sudden changes in Greylag numbers is unclear. The lake is very small for so many geese and easily disturbed. Even so, geese did not start coming in after dark, as they now do at other sites in the north-east. There are no known alternative roost sites nearby, although 3000 Greylag were seen at Fyvie Lake in November 1979 (Owen et a/. 1986), which appears to predate the large numbers at Haddo House. They have not been seen subsequently at Fyvie.

Donside and Deeside

Loch of Skene and Kemnay The Loch of Skene is large (118 ha), shallow and rather exposed on the low ground between lower Deeside and lower Donside. There was an autumn peak, low numbers in mid-winter depending on snow cover, and a spring peak. The Loch of Skene was sub- ject to regular wildfowling. There were no sites close by which were suitable as permanent roosts but there were several semi-permanent marshes and pools in the Don valley to the north. These were probably used occasionally and may explain some of the fluctuations in numbers at Skene.

When the November counts of Greylag were started in 1960 there were several hundred roosting on the River Don near Kemnay and none used the Loch of Skene. The first record at Skene was of 150 in November 1964 and numbers in- creased to 2700 in 1974. During this period numbers at Kemnay were very stable at around 200 birds. There were no further records of geese roosting at Kemnay until late November 1987 when c.1500 used the site for about a week (G. Wright pers. comm.).

Fig. 5a shows the November counts for Loch of Skene since 1960. Both the November 1983 and November 1984 counts were thought to be in- complete; in 1983 4000 were present a week either side of the count date and in 1984 a flood in the Don valley probably attracted many birds. After the initial stabilization at c.2500 birds in the mid-1970s there was an increase to about 4000 birds between 1978 and 1984, then a further in- crease in 1985. The March counts between 1982 and 1986 varied from 3210-5060 birds which was one of the largest flocks of Greylag in Britain at this season.

In the early years of the counts small numbers of Pinkfeet also occurred in nine autumns with a maximum of 350 in November 1969. Very few have been recorded since. At Kem- nay up to 300 Pinkfeet were feeding in riverside fields on the wildfowl count days between the 1967/68 and 1976/77 winters but none have been recorded since.

Up to the 1985/86 winter limited data sug- gested an autumn peak of 4000-6500 Greylag at Skene in late October/early November with a slight decrease by mid-November. The largest autumn number was 6540 on 20 October 1981, and 8850 seen on 19 January 1985 was exceptional at that time. There were usually less than 2500 in mid-winter, and a spring peak of up to 5000 birds in late March/early April. The numbers us- ing the Loch of Skene in the 1985/86 winter are

1988 Geese in NE Scotland 55 a 10 a

>

G)

5

=

5

=

£ ot —_—_

1960 65 70 75 80

Thousands of Greylag

Oct Nov Dec Jan Feb

85 15 Cc fo.) & ra (5 «10 wo) wo Oo e ise) 4 rs) < @) Ee Mar Apr 1975 80 85

FIGURE 5. a) November counts of Greylag at Loch of Skene b) Counts of Greylag roosting at Loch of Skene during the 1985/86 winter c) November counts of Greylag at the Dinnet Lochs. The arrows indicate snowfalls and the bars periods with 10 cm or more snow cover.

shown in Fig. 5b. This winter was atypical as the numbers were the highest ever recorded here. The autumn peak of almost 9000 birds was the largest by several thousand and quite large numbers were present in mid-winter in spite of several snowfalls, with 8950 present on 5 January.

Corby Loch Corby Loch is a small loch (12 ha) near the coast just to the north of Aberdeen. It is unclear when Greylag first started to use this site and it was not counted systematically until the 1983/84 winter. It was known before then that several hundred geese often fed in the fields just to the north and south of the loch and occasionally when large numbers were present they flighted out to the lower Don valley. However, records in the mid-1980s suggested that substantial numbers of geese used the loch, particularly in spring. Early in the winter birds were shot and the autumn flocks tended to be smaller than those in spring.

The numbers of Greylag found at Corby Loch on the national counts since 1983 have rang- ed from 0-2000 (November) and from 800-2400 (March). Counts of 1600 on 9 November 1975 and

1150 on 11 November 1979 at the loch during the day were not included in the national counts for those years although the geese were almost cer- tainly roosting there. Small numbers of Pinkfeet were occasionally found at Corby Loch, the largest being 140 on 22 November 1981.

The Dinnet Lochs and middle Deeside Small numbers of Greylag began using flooded fields at Auchlossan, the site of an old loch in middle Deeside, during the spring in the late 1960s. At this time they probably roosted on the flood itself, or on the River Dee at Ballogie, a site which is still occasionally used by up to 1500 geese. Small numbers also appeared at the Dinnet Lochs in the 1960s, particularly in the autumn, but they did not remain for long and it was not until the mid-1970s that the site was used regularly and numbers then increased rapidly. It is now one of the most important autumn concentrations of Greylag in Britain.

The Dinnet Lochs were first counted in November in 1974 and have been checked every year since. The lochs were included in the Muir

56 =M.V. Bell et al. SB 15 (2)

Uz 2 10 10

8 8

6 6

yi (eset naar A Me

Oo : ‘My SOR : ® ees? O O O N D

| 1983/84

Thousands of Greylag

Onhoh News oad) Fre MamA

1985/86

J Fini A

1984/85 16 14 2 10 8 6 4 2

a * O N D J F MseeA 1986/87

FIGURE 6. The numbers of Greylag roosting at the Dinnet Lochs in the 1983/84 — 1986/87 winters.

@ = roost counts, o = field counts. The arrows represent snowfalls and the bars periods with 10 cm or more snow cover. Records of snowfall and snow cover were not available for the 1986/87 winter, and no field counts were undertaken in spring 1987.

of Dinnet NNR in 1977 and since the 1983/84 winter the warden has done regular counts. There was no shooting of ducks or geese at the Dinnet Lochs. The November counts at Dinnet show a series of sudden increases interspersed by a few years when numbers were stable (Fig. Sc). Bet- ween 1974 and 1977 800-1700 were recorded, bet- ween 1978 and 1981 there were 3500-4500 and there was a further large increase in 1982. The pattern of the autumn peak also changed slight- ly over the years. In the late 1970s numbers in- creased gradually through late October and November and were often still increasing when the first snowfall drove the geese south in the se- cond half of November or in December. Very few were then seen until the spring. However, in the autumns of 1983 and 1984 they increased rapid- ly to peak in early November, then quickly decreased in the absence of adverse weather (Fig. 6). In 1985 the arrival was rather early with over

3000 on 12 October, 16,000 by 3 November and a peak of 19,900 on 26 November. Over the following three days snow caused all the geese to leave but unlike previous winters within a week large numbers were returning and a second peak of 12,300 occurred on 18 December. After fur- ther snow only small numbers remained through January and February, but there was a definite tendency for geese to remain here throughout the 1984/85 and 1985/86 winters. The spring peak of 4600 in 1986 was the largest on record, but spring counts were difficult because the geese did not always roost on the lochs, and field counts were usually the most reliable at this time (Fig. 6). In late October 1986 over 13,000 were present and the peak count of 15,200 occurred on 24 November. In spring 1987 Greylag roosted at the lochs on only six nights from early to mid-March although they were present in the area until mid- April.

}

1988 Geese in NE Scotland 57

TABLE 3. Summaries of the November counts showing the north-east Scotland and national totals of Pinkfeet and Greylag in thousands, and the regional total as a percentage of the British total.

PINKFEET GREYLAG Britain NE Scotland % of British Britain NE Scotland % of British

total total 1957-59 35.4 ts 4.8 1960-64 58.1 2.6 4.5 Bos 2.8 79 1965-69 69.9 13.1 18.7 55.5 6.4 TS 1970-74 76.2 Aes 9.6 68.4 9.9 14.5 1975-79 74.2 lez 15.8 67.8 126 18.6 1980-84 96.9 12.4 12.8 88.8 20.2 227 1985-86 129.3 25.5 19.7 T1222 25.1 22.4

Note. Greylag were first counted in 1961 at the Loch of Strathbeg, Slains Lochs, Loch of Skene and Kemnay. The Dinnet Lochs were first counted in 1974, Fedderate Reservoir in 1977, Haddo House Lake in 1981 and Corby Loch in 1983. Kemnay was not counted after 1976. Data summarized

from Boyd & Ogilvie 1969, 1972; Ogilvie & Boyd 1976; Wildfowl Trust Annual Reports.

There were a few other places on south Deeside where feeding Greylag were recorded, and some may have been birds stopping for a short period before crossing the hills. On 18 March 1986 Dr A.G. Knox found a flock of 544 Greylag at a site on south Deeside where geese had not previously been recorded which illustrated the dif- ficulty of attempting complete counts of Greylag in spring.

In autumn small flocks of Pinkfeet flew over Deeside and skeins sometimes stopped at the Dinnet Lochs for a few hours, but they rarely fed in the area. However, in spring small numbers, usually under 80 birds, regularly occurred for a few weeks in the Howe of Tarland and in some years many more were present. There were 780 there on 8 April 1979, 630 on 21 April 1985 and 890 on 20 April 1986. In spring 1987 up to 900 Pinkfeet were present in the Howe of Tarland from mid-February to April. Up to 600 roosted at the Dinnet Lochs on seven nights between 16 February and 21 April.

Discussion

North-east Scotland currently holds some very large concentrations of both Pinkfeet and Greylag, and the numbers of both species wintering in the area have increased greatly over the last 25 years. The November

counts between the early 1960s and mid-1980s showed a tenfold increase in Pinkfeet numbers in north-east Scotland against a four to fivefold increase nationally while Greylag also increased eight to tenfold in north-east Scotland but only two to threefold nationally (Table 3). The area regularly holds 15-30% of the British total of Pinkfeet and 20-35% of the Greylag. The regional distributions of the two species in autumn have previously been analysed (Boyd & Ogilvie 1969, 1972; Ogilvie & Boyd 1976) with the latter paper predicting in- creases in Aberdeenshire in both absolute and relative terms for Pinkfeet and smaller increases for Greylag. This has since occur- red though Greylag have shown the largest gains (Table 3). The increase for Pinkfeet has been rather erratic, largely due to a tem- porary rise in the late 1960s when counts at Meikle Loch were especialy high (see Fig. 3). Greylag have shown a steadier increase since 1960, the highest November total be- ing 29,300 in 1982 which was 36.6% of the national total. They have increased by oc- cupying five new roosts while the Pinkfeet are still concentrated at the two traditional sites. In recent autumns the largest numbers of Pinkfeet occurred in mid-October when

58 M.V. Bell et al.

SB 15 (2)

Strathbeg

Slains

Grass Potatoes “=! Oats

Barley

Wheat

FIGURE 7. The proportion of different crop types as a percentage of the total cultivated area from selected parishes close to each of the main roosts. Agricultural statistics were obtained from the Scottish Records Office for the years 1960, 1965, 1970, 1975, 1978, 1980, 1982 and 1984 for the parishes of Lonmay, Crimond and St Fergus (Strathbeg), Slains, Logie-Buchan and Foveran (Slains), Methlick and Meldrum (Haddo), Kinellar and Skene (Skene), and Lumphanan and Leochel-Cushnie (Dinnet).

Loch of Strathbeg and Meikle Loch have together held up to 40,000 birds. North-east Scotland also holds large numbers of both species in the spring with up to 24,000 Greylag and 30,000 Pinkfeet present in March and April respectively.

There were no changes at the roost sites to make them more attractive to geese in this period but there were substantial changes in the farming regimes on the feeding areas. Fig. 7 shows how the cropping pattern has changed on the farmland adjacent to the main roosts between 1960 and 1984.

The data on crop areas are presented for each crop type as a percentage of the total cultivated area (which includes grass but not rough grazing) for the parishes rele- vant to each roost (Fig. 7). Between 80% and 90% of the grass was less than seven years old. During this period grass (45-65%) and cereals (25-50%) were the main crops but there were major changes in the cereal regime. Before 1960 most of the cereal

acreage was oats but over the following 10-15 years barley almost completely replac- ed the oats near the coast while the change was slower further inland (Fig. 7). The small area of wheat remained fairly steady bet- ween 1960 and 1980 but since then increased markedly to about 10% of total crop area in the Ythan valley by 1984. There has also been a major change to autumn sown winter cereals in recent years. Subjective impres- sions were that winter cereals were rarely sown before